- Email: [email protected]
The determination of platinum in biological tissue by instrumental neutron activation analysis
Appl. Radiat. Isot. Vol. 42, No. 8, pp. 775-776, 1991 Int. J. Radiat. Appl. Instrum. Part A
© Pergamon Press pie 1991. Printed in Great Britain 0883-2889/91 $3.00 + 0.00
The Determination of Platinum in Biological Tissue by Instrumental Neutron Activation Analysis G. W. HAVERLAND* and L. I. WIEBE University of Alberta SLOWPOKE Facility, Edmonton, Alberta, Canada T6G 2N8 (Received 27 December 1989; in revised form 23 January 1991) Neutron activation analysis was used to determine platinum in the tissues of rats that had received doses of the tumor chemotherapy agent cis-Pt. Two elements causing spectral interference were identified as 475c from 46Ca, a natural component of tissue, and tin (t17mSn), an element introduced inadvertently during necropsy and sample work-up. A protocol was devised to correct for interference, using emission lines from 47Ca and tt3Sn, respectively, with appropriate consideration for relative detector efficiencies, decay times and half-lives. Introduction
One of the methods available for the quantitative estimation of platinum in a biological matrix is neutron activation analysis (NAA) (Sykes, 1988). A sensitive nuclear reaction for NAA of Pt is based on the measurement of 198pt, which has a natural abundance of 7.2%. The activation reaction, 198pt (n, 7) 199pt gives rise to the radioactive daughter 199Au which in turn decays to t99Hg. Two gamma rays from the decay of 199Au( 1 5 8 keV [37%] and 208 keV [8.4%]) (Browne and Firestone, 1986) can be measured directly in samples which have no matrix interferences. However, their quantification in certain matrices is complicated by the presence of spectral interferences. The presence of calcium or tin, for example, both give rise to interferences with the 158 keV photopeak due to the emission of 159 keV gamma rays by 475c (daughter of 47Ca) and 117mSn,respectively. Corrections for these interferences as part of a program to assay for Pt in an experimental tumor model are now described.
Selected tissues were collected upon necropsy, from rats having received injections of the antitumor drug cis-Pt. These samples, which have unknown Pt content, were frozen and stored at - 1 5 ° C until analysis was undertaken. In preparation for analysis, the samples were allowed to thaw slightly and then transferred to clean 1.5 mL polyethylene vials. The samples were dried in still air at ca 60°C and then compacted with a clean glass rod, after which the vials were heat sealed. The samples and standards were irradiated for 4 h at a nominal neutron flux of 1 x 1012n/cm2s, allowed to decay for 4 to 17 days and then counted for 6 h (live time). Samples were counted at a fixed distance (detector face to sample centroid) of I cm, unless the detector dead-time was greater than 5%, in which case a distance of 2cm was used to reduce pulse pile up and coincidence summing. Measurements were taken using a Canberra C413R coaxial Ge(Li) detector (11% efficiency, 1.88 keV FWHM) together with a Canberra Series 80 multichannel analyzer. The spectra from the samples and the standards were analyzed using the Nuclear Data (ND) standard analysis program (Nuclear Data, 1977).
Results and Discussion
A preliminary study showed that the best sensitivity for Pt occurred when the sample decay time was between 7 and 14 days. The peak searches of spectra from both the clean paper standards and the mouse tissue standards identified the ~99Au 158 keV photopeak. These peak areas were statistically similar (P < 0.1) for standards with the same nominal Pt content, thereby establishing that both sets of standards were representative of a single distribution. These data were used to prepare a calibration line in which counts are plotted as a function of #g Pt (Fig. 1). For the biological samples containing unknown amounts of Pt, however, this initial analysis using only the 158 keV photopeak produced data inconsistent with the experimental protocol in that some samples contained an appreciable 158 keV photopeak when no cis-Pt had been administered to the animal. On the basis of spectrum analysis, this unexpected photopeak was attributed to the presence of
250 r~0
200 0
150 c-
Experimental
Two types of Pt reference standards were prepared, one using commercial paper previously treated to reduce the Na, AI, Mn and CI content of the matrix (Haverland, 1988) and the second using homogenized murine tissue. Both types of standards were prepared to have 1, 2, 4 or 8 #g (nominal) of Pt per vial. Platinum used for the reference standards was obtained from a diluted commercial atomic absorption standard (BDH Platinum standard, 1000ppm, diluted to 1/~g/100#L). One set of standards consisted of five Ptspiked tissue standards and the other set consisted of 9 Pt-spiked paper standards. *Author for correspondence. 775
o o "o
I_ o
100 -
#
50-
o
0
I
I
1
2
I
I
I
3 4 5 6 7 M i c r o g r a m s p la t in u m
I
I
I
I
8
9
Fig. 1. Calibration curve for Pt derived from NAA of known Pt samples on a matrix of clean paper or animal tissue. Samples were analyzed at either 1 or 2 era counting geometries and counts were then converted to counts at I cm to obtain corrected counts.
776
Technical Note
calcium and/or tin in some of the samples; both of these elements have isotopes which activate to produce spectral interferences, with corresponding half-lives of 4.536 days and 115.1 days, respectively. For calcium, the interference was attributed to neutron activation of 46Ca which has a natural abundance of 0.0033% (Browne and Firestone, 1986): 46Ca(n, 7 ) 4 7 C a
,, # ~47Sc 4.536 days 1297keV 75%
~3.341 days 159keV 68%
47Ti'
The gamma spectra from three unknown biological samples contained the 1297 keV photopeak. It was known from the experimental protocol that two of these samples came from animals not treated with cis-Pt and therefore an appropriate adjustment to the 158 keV photopeak area was required. The correction to the 158 keV photopeak area was based on the assumption that at the time of measurement the 47Scwas almost in secular equilibrium and by compensation for the known relative efficiency of detection for the gamma ray energies in question. The quantity of 47Ca in the sample was estimated from the 1297keV photopeak. The 158keV photopeak was in fact removed from the two spectra by making the calcium correction in this manner; the third sample had the 158 keV photopeak area reduced by 17% by applying this correction factor. For tin, the interference from the decay of 1lTmZnproduced by the activation of ~J6Sn (natural abundance 14.24%) (Browne and Firestone, 1986) was determined by quantitating the amount of ~2Sn (natural abundance 0.95%) present: it3in N2Sn(n' ' y ) l l 3 S n liSAE.C. days I 392 keV 64%
U6Sn(n, ~ , ) N T m S n ~
13.6 days 159keV 85%
117Sn
The tin correction was calculated by correcting the H3Sn 392 keV photopeak counts to end of bombardment (EOB), multiplying these corrected counts by the expected ratio (0.145) of HTmSn 159 keV to H3Sn 392 keV counts at EOB, and finally correcting these counts back to the actual counting time. The ratio of detection efficiencies (392 keV vs 159 keV) for the detector used is ca 0.45. Consequently, the number of counts at EOB is ca 15 times greater for 159 keV than for 392 keV (.~ 1/0.145.0.45) for a pure tin sample. The ratio of 117mSn 159 keV to t t3Sn 392 keV counts was obtained by correcting to end of bombardment, the counts observed in 18 h after ca 17 days decay. Spectra from 22 unknown biological samples required adjustment of the 158 keV photopeak area. Ten of these samples were not expected to contain Pt and eight of these ten samples were given corrections > 75%. Of the remaining 12 samples, which were expected to contain Pt because the experimental tumor bearing animal was treated with cis.Pt, adjustments ranged from 6 to 100%. These adjustments were applied to seven samples taken from animals which received low doses of cis-Pt and in two cases, where both samples were derived from the same animal, corrections of 95 and 100% were required for tumor and serum samples, respectively. For the remaining samples from animals in the high Pt dose group, the correction factors ranged from 6 to 61%.
Even with the corrections for calcium and tin to the 158 keV photopeak areas, the final matrix of results showed some large inconsistencies: samples counted multiple times and samples in the same treatment group sometimes had widely varying estimates of Pt concentration. This variation was thought to be due, at least in part, to the peak search algorithm. About ~ of the spectra were analyzed manually in an attempt to improve this. Peak areas were determined by setting the peak boundaries where they (globally) minimized the fractional error of the net peak area. As with the ND routines, the background was assumed to be linear under the photopeak, although this is known not to be the case (Jorch and Campbell, 1977). The manual peak search results for a subset of the spectra that appeared to have consistent estimates of Pt concentration were checked against the results obtained by the ND routines to test for bias. The results were found to be substantially the same, well within the expected error in photopeak counts.
Summary Platinum in biological tissue was determined at the ppm level using instrumental NAA with the 198pt (n,~) 199pt activation reaction, followed by radiometry on the 159 keV photopeak of the radioactive daughter 199Au. Tin and calcium were determined to be interferences in this analysis, the former suspected to be derived from tin solder used in the area were animal samples were taken during necropsy of cis-Pt-treated animals, and the latter as a natural component in relatively high concentration in some tissues. Quantification of the interfering elements, using other characteristic photopeaks, provided a basis for adjustment of the 158 keV photopeak area, thereby greatly improving the level of confidence in analytical data from samples of unknown platinum content.
Acknowledgements--The University of Alberta SLOWPOKE Facility is supported in part by the Alberta Heritage Foundation for Medical Research (Grant No. SIF 6882). The authors wish to thank Drs Brian C. Lentle and Gerd Theyer, Vancouver General Hospital, for providing biological tissues from cis-Pt-treated tumor bearing mice for this investigation.
References Browne E. and Firestone R. B. (1986) In Table of Radioactive Isotopes (ed. Shirley V. S.), pp. 198-1, 199-I, 47-1, 113-3, 117-5. Wiley, New York. Haverland G. W. (1988) In University of Alberta SLOWPOKE Facility Annual Report, November 1988 (eds Wiebe L. I. and Ford P.), p. R25. University of Alberta SLOWPOKE Facility, Edmonton. Jorch H. H. and Campbell J. L. (1977) On the analytic fitting of full energy peaks from Ge(Li) and Si(Li) photon detectors. Nucl. Instr. Meth. 143, 551-559. Nuclear Data, Inc. (1977) In ND6600 Basic Physics Application Package Program Algorithms. Nuclear Data, Inc., Schaumburg. Sykes T. R. (1988) In Quantitative Trace Analysis of Biological Materials (eds McKenzie H. A. and Smythe L. E.), p. 659. Elsevier, New York.
© Pergamon Press pie 1991. Printed in Great Britain 0883-2889/91 $3.00 + 0.00
The Determination of Platinum in Biological Tissue by Instrumental Neutron Activation Analysis G. W. HAVERLAND* and L. I. WIEBE University of Alberta SLOWPOKE Facility, Edmonton, Alberta, Canada T6G 2N8 (Received 27 December 1989; in revised form 23 January 1991) Neutron activation analysis was used to determine platinum in the tissues of rats that had received doses of the tumor chemotherapy agent cis-Pt. Two elements causing spectral interference were identified as 475c from 46Ca, a natural component of tissue, and tin (t17mSn), an element introduced inadvertently during necropsy and sample work-up. A protocol was devised to correct for interference, using emission lines from 47Ca and tt3Sn, respectively, with appropriate consideration for relative detector efficiencies, decay times and half-lives. Introduction
One of the methods available for the quantitative estimation of platinum in a biological matrix is neutron activation analysis (NAA) (Sykes, 1988). A sensitive nuclear reaction for NAA of Pt is based on the measurement of 198pt, which has a natural abundance of 7.2%. The activation reaction, 198pt (n, 7) 199pt gives rise to the radioactive daughter 199Au which in turn decays to t99Hg. Two gamma rays from the decay of 199Au( 1 5 8 keV [37%] and 208 keV [8.4%]) (Browne and Firestone, 1986) can be measured directly in samples which have no matrix interferences. However, their quantification in certain matrices is complicated by the presence of spectral interferences. The presence of calcium or tin, for example, both give rise to interferences with the 158 keV photopeak due to the emission of 159 keV gamma rays by 475c (daughter of 47Ca) and 117mSn,respectively. Corrections for these interferences as part of a program to assay for Pt in an experimental tumor model are now described.
Selected tissues were collected upon necropsy, from rats having received injections of the antitumor drug cis-Pt. These samples, which have unknown Pt content, were frozen and stored at - 1 5 ° C until analysis was undertaken. In preparation for analysis, the samples were allowed to thaw slightly and then transferred to clean 1.5 mL polyethylene vials. The samples were dried in still air at ca 60°C and then compacted with a clean glass rod, after which the vials were heat sealed. The samples and standards were irradiated for 4 h at a nominal neutron flux of 1 x 1012n/cm2s, allowed to decay for 4 to 17 days and then counted for 6 h (live time). Samples were counted at a fixed distance (detector face to sample centroid) of I cm, unless the detector dead-time was greater than 5%, in which case a distance of 2cm was used to reduce pulse pile up and coincidence summing. Measurements were taken using a Canberra C413R coaxial Ge(Li) detector (11% efficiency, 1.88 keV FWHM) together with a Canberra Series 80 multichannel analyzer. The spectra from the samples and the standards were analyzed using the Nuclear Data (ND) standard analysis program (Nuclear Data, 1977).
Results and Discussion
A preliminary study showed that the best sensitivity for Pt occurred when the sample decay time was between 7 and 14 days. The peak searches of spectra from both the clean paper standards and the mouse tissue standards identified the ~99Au 158 keV photopeak. These peak areas were statistically similar (P < 0.1) for standards with the same nominal Pt content, thereby establishing that both sets of standards were representative of a single distribution. These data were used to prepare a calibration line in which counts are plotted as a function of #g Pt (Fig. 1). For the biological samples containing unknown amounts of Pt, however, this initial analysis using only the 158 keV photopeak produced data inconsistent with the experimental protocol in that some samples contained an appreciable 158 keV photopeak when no cis-Pt had been administered to the animal. On the basis of spectrum analysis, this unexpected photopeak was attributed to the presence of
250 r~0
200 0
150 c-
Experimental
Two types of Pt reference standards were prepared, one using commercial paper previously treated to reduce the Na, AI, Mn and CI content of the matrix (Haverland, 1988) and the second using homogenized murine tissue. Both types of standards were prepared to have 1, 2, 4 or 8 #g (nominal) of Pt per vial. Platinum used for the reference standards was obtained from a diluted commercial atomic absorption standard (BDH Platinum standard, 1000ppm, diluted to 1/~g/100#L). One set of standards consisted of five Ptspiked tissue standards and the other set consisted of 9 Pt-spiked paper standards. *Author for correspondence. 775
o o "o
I_ o
100 -
#
50-
o
0
I
I
1
2
I
I
I
3 4 5 6 7 M i c r o g r a m s p la t in u m
I
I
I
I
8
9
Fig. 1. Calibration curve for Pt derived from NAA of known Pt samples on a matrix of clean paper or animal tissue. Samples were analyzed at either 1 or 2 era counting geometries and counts were then converted to counts at I cm to obtain corrected counts.
776
Technical Note
calcium and/or tin in some of the samples; both of these elements have isotopes which activate to produce spectral interferences, with corresponding half-lives of 4.536 days and 115.1 days, respectively. For calcium, the interference was attributed to neutron activation of 46Ca which has a natural abundance of 0.0033% (Browne and Firestone, 1986): 46Ca(n, 7 ) 4 7 C a
,, # ~47Sc 4.536 days 1297keV 75%
~3.341 days 159keV 68%
47Ti'
The gamma spectra from three unknown biological samples contained the 1297 keV photopeak. It was known from the experimental protocol that two of these samples came from animals not treated with cis-Pt and therefore an appropriate adjustment to the 158 keV photopeak area was required. The correction to the 158 keV photopeak area was based on the assumption that at the time of measurement the 47Scwas almost in secular equilibrium and by compensation for the known relative efficiency of detection for the gamma ray energies in question. The quantity of 47Ca in the sample was estimated from the 1297keV photopeak. The 158keV photopeak was in fact removed from the two spectra by making the calcium correction in this manner; the third sample had the 158 keV photopeak area reduced by 17% by applying this correction factor. For tin, the interference from the decay of 1lTmZnproduced by the activation of ~J6Sn (natural abundance 14.24%) (Browne and Firestone, 1986) was determined by quantitating the amount of ~2Sn (natural abundance 0.95%) present: it3in N2Sn(n' ' y ) l l 3 S n liSAE.C. days I 392 keV 64%
U6Sn(n, ~ , ) N T m S n ~
13.6 days 159keV 85%
117Sn
The tin correction was calculated by correcting the H3Sn 392 keV photopeak counts to end of bombardment (EOB), multiplying these corrected counts by the expected ratio (0.145) of HTmSn 159 keV to H3Sn 392 keV counts at EOB, and finally correcting these counts back to the actual counting time. The ratio of detection efficiencies (392 keV vs 159 keV) for the detector used is ca 0.45. Consequently, the number of counts at EOB is ca 15 times greater for 159 keV than for 392 keV (.~ 1/0.145.0.45) for a pure tin sample. The ratio of 117mSn 159 keV to t t3Sn 392 keV counts was obtained by correcting to end of bombardment, the counts observed in 18 h after ca 17 days decay. Spectra from 22 unknown biological samples required adjustment of the 158 keV photopeak area. Ten of these samples were not expected to contain Pt and eight of these ten samples were given corrections > 75%. Of the remaining 12 samples, which were expected to contain Pt because the experimental tumor bearing animal was treated with cis.Pt, adjustments ranged from 6 to 100%. These adjustments were applied to seven samples taken from animals which received low doses of cis-Pt and in two cases, where both samples were derived from the same animal, corrections of 95 and 100% were required for tumor and serum samples, respectively. For the remaining samples from animals in the high Pt dose group, the correction factors ranged from 6 to 61%.
Even with the corrections for calcium and tin to the 158 keV photopeak areas, the final matrix of results showed some large inconsistencies: samples counted multiple times and samples in the same treatment group sometimes had widely varying estimates of Pt concentration. This variation was thought to be due, at least in part, to the peak search algorithm. About ~ of the spectra were analyzed manually in an attempt to improve this. Peak areas were determined by setting the peak boundaries where they (globally) minimized the fractional error of the net peak area. As with the ND routines, the background was assumed to be linear under the photopeak, although this is known not to be the case (Jorch and Campbell, 1977). The manual peak search results for a subset of the spectra that appeared to have consistent estimates of Pt concentration were checked against the results obtained by the ND routines to test for bias. The results were found to be substantially the same, well within the expected error in photopeak counts.
Summary Platinum in biological tissue was determined at the ppm level using instrumental NAA with the 198pt (n,~) 199pt activation reaction, followed by radiometry on the 159 keV photopeak of the radioactive daughter 199Au. Tin and calcium were determined to be interferences in this analysis, the former suspected to be derived from tin solder used in the area were animal samples were taken during necropsy of cis-Pt-treated animals, and the latter as a natural component in relatively high concentration in some tissues. Quantification of the interfering elements, using other characteristic photopeaks, provided a basis for adjustment of the 158 keV photopeak area, thereby greatly improving the level of confidence in analytical data from samples of unknown platinum content.
Acknowledgements--The University of Alberta SLOWPOKE Facility is supported in part by the Alberta Heritage Foundation for Medical Research (Grant No. SIF 6882). The authors wish to thank Drs Brian C. Lentle and Gerd Theyer, Vancouver General Hospital, for providing biological tissues from cis-Pt-treated tumor bearing mice for this investigation.
References Browne E. and Firestone R. B. (1986) In Table of Radioactive Isotopes (ed. Shirley V. S.), pp. 198-1, 199-I, 47-1, 113-3, 117-5. Wiley, New York. Haverland G. W. (1988) In University of Alberta SLOWPOKE Facility Annual Report, November 1988 (eds Wiebe L. I. and Ford P.), p. R25. University of Alberta SLOWPOKE Facility, Edmonton. Jorch H. H. and Campbell J. L. (1977) On the analytic fitting of full energy peaks from Ge(Li) and Si(Li) photon detectors. Nucl. Instr. Meth. 143, 551-559. Nuclear Data, Inc. (1977) In ND6600 Basic Physics Application Package Program Algorithms. Nuclear Data, Inc., Schaumburg. Sykes T. R. (1988) In Quantitative Trace Analysis of Biological Materials (eds McKenzie H. A. and Smythe L. E.), p. 659. Elsevier, New York.