J. Trace Elements Med. BioI. Vol. 9, pp. 193-199 (l995) Br, Rb, Zn, Fe, Se and K in Blood of Colorectal Patients by INAA and PIXE C. SHENBERG 1, H. FELDSTEIW, R. CORNELIS, L. MEES, J. VERSIECK*, L.VANBALLENBERGHE *, J. CAFMEYER AND W. MAENHAUT Laboratory for Analytical Chemistry, Institute for Nuclear Sciences, Proeftuinstraat 86, B-9000, Gent, Belgium *Department of Internal Medicine, Division of Gastroenterology, University Hospital, De Pintelaan 185, B-9000 Gent, Belgium (Received June/August 1995) Summary A preliminary study was conducted on blood samples and blood fractions of 11 colorectal patients and 10 healthy subjects (controls) in Belgium, in order to determine the concentration of some vital trace elements. Two non destructiv,y analytical methods were used for the determination: INAA and PIXE. The agreement between PIXE and INAA was within about ± 10% for plasma, but for Rb, Se and Fe in whole blood and red cells a difference of ± 20% was noted; part of the discrepancy may be due to self absorption problems in PIXE, and for Rb, spectral interferences also may have contributed. The precision of the INAA method for the elements studied was found to be ± 3% for whole blood and red cells and ± 5 % for plasma; the accuracy for Br, Rb and Zn was better than ± 10% and ± 17% for Se. The ratios of the concentrations in whole blood to red cells and whole blood to plasma were not significantly different for normals and cancer cases and, therefore, in future studies analysis of whole blood only may be sufficient. The mean values for Br, Rb, BrlRb ratio, K, Fe and Se were significantly lower for cancer cases than for healthy individuals, and this might be applicable as an additional parameter for differentiating normals from malignant cases. Keywords: Trace elements, Br, K, Fe, Rb, Se, blood, colorectal cancer, INAA, PIXE. Introduction An increasing number of experimental and epidemiological studies suggest that trace elements play an essential role in the human body 0-3). A number of sources indicate a relationship between trace elements and the incidence of cancer in the human population (4-8). Very little is known about Br and Rb status and their possible role in human pathology. The implication of bromine and bromine-containing compounds was discussed in relation to ITo whom correspondence should be addressed at the Soreq Nuclear Research Center, Yavne-81800, Israel. Temporarily on sabbatical leave in the Laboratory of Analytical Chemistry, Gent. © 1995 by Gustav Fischer Verlag Stuttgart· Jena . New York the stupor of uremic patients (9). The role of Rb in the differentiation of mononuclear cells (0) and several works relating to Rb in psychiatry, were reported (lIB). In previous publications 04-16) we reported on trace element relative concentrations in normal and malignant tissues, determined by high sensitivity X-Ray Fluorescence (XRF). Statistically significant differences in the concentrations of alkali and halogen elements were found in the normal and malignant colon tissue of 12 patients suffering from colorectal cancer (CRC); in tumorous tissue the levels of K and Rb were increased, whereas the levels of Br and CI were decreased with respect to normal tissue (4). In an experiment with a relatively large number of mice inoculated with lung carcinoma, it 194 C. Shenberg, H. Feldstein, R. Comelis, L. Mees, J. Versieck, L.VanbaIlenberghe, J. Cafmeyer and W. Maenhaut has been shown that a significant difference exists between the Br/Rb ratio in nonnal and malignant kidney tissues of these animals (15). Recently, the distlibution of BrKa/RbKa peak intensity ratio was studied in human whole blood samples obtained from 61 CRC patients and 124 healthy subjects (controls) in Israel (16). Based on the results already obtained, we decided to extend our previous experiments to include populations from other countries, in order to detennine whether our findings are universal. The aim of the present work was to conduct a preliminary study on blood samples of CRC patients in Gent, Belgium. Two analytical methods were applied - Instrumental Neutron Activation Analysis (INAA) and Particle-Induced X-ray Emission (PIXE). The PIXE method is much faster, but requires proper correction for matrix effects and can only measure elements down to 0.1 I-lg/g dry weight. Both analytical methods detennine many elements simultaneously and therefore the concentrations of Cu, Se, Zn, Fe and K in the blood of healthy subjects and CRC patients will be reported as well. Finally, we aimed to obtain infonnation regarding the distribution of Br, Rb and other elements between the blood fractions (plasma and red cells) and to report the data in the most suitable way. Table 1. Characteristic data of the individuals studied Total number Number of males Number of females Mean age (y) Median age (y) Range of age Cancer patients Controls II LO 6 5 78 80 56 - 95 5 5 45 45.5 32 - 57 Characteristic data of the cancer patients N° Age Sex Description of the disease state l' 80 M 2 65 F 3' 56 F 4 5 87 76 M F 6 7 88 95 F 8 80 M 9 66 M Materials and Methods LO 80 M Subjects and samples 11 85 F Moderately differentiated adenomacarcinoma of the sigmoid. Liver metastases. Laser therapy, chemotherapy Resection of adenomacarcinoma of the sigmoid in 1989. Solitary liver metastases Resection of adenomacarcinoma of the rectasigmoid. Post-operative chemo- and radiotherapy Carcinoma of the rectum. Laser therapy Adenomacarcinoma of the rectum. Diagnosis in 1992. Laser therapy Malign villous tumour of the sigmoid Moderately differentiated adenocarcinoma of the rectum. Diagnosis in 1992. Endoscopic restriction.Lasertherapy. Perirectal invasion Adenomacarcinoma of the rectum. Infiltration of the perirectal tract. Sampling before start of therapy Moderately differentiated adenomacarcinoma of the caecum Well differentiated adenomacarcinoma in tubulo-villous tumour. Endoscopic resection. Laser therapy Carcinoma of the rectosigmoid. Laser therapy A total of 11 CRC patients (two patients were sampled twice with a time interval of several weeks) treated at the University Hospital, Gent, Belgium, participated in the present study. 10 apparently healthy workers from the Hospital and from the Institute for Nuclear Sciences in Gent served as controls. The characteristic data of the subjects studied is given in Table 1. Blood samples from patients and controls were collected at the Department of Internal Medicine, Division of Gastroenterology of the University Hospital. 1-2 drops of stable pure heparine anticoagulant solution (Anticlot-Clinton Biochimie, B. Pat: 5129) were added to 5 ml blood samples collected in plastic tubes. In order to avoid potentially detrimental changes, the samples were transferred for further treatment to the Institute for Nuclear Sciences within a few hours of collection. All processing was done under clean laboratory conditions; all materials used were carefully cleaned by rinsing with milli-Q water and s ubboiled mineral acids. Upon arrival in the clean room, part of the blood fro m each sample was separated into plasma and red cells by centrifugation (twice 10 min. 3000 rpm, 1200 g). Whole blood samples M 'Patients 1 and 3 were sampled twice with time intervals of 6 and 8 weeks respectively. before centrifugation, red cells and plasma samples after separation were carefully homogenized and divided into two portions; one for INAA and one for PIXE analysis. INAA 1-2 ml fractions of whole blood, red cells and plasma were weighed in poiyethylene irradiation vials, deep frozen and freeze-dried (starting temperature appro x -35°C) for 15 h. Neutron flux monitors prepared from iron foil were sandwiched between the samples in irradiation containers. 12 different samples (duplicates of all fractions ) were prepared for each case; 6 samples for Br and K detennination (3 h irradiation) and 6 additional samples for Rb, Zn, Fe and Se detennination (14-35 h irradiation). The samples were irradiated with a thennal neutron flux of 1.1 x 10 12 n.cm·2 S·I in the Thetis reactor at the Institute for Nuclear Sciences, Gent and the gamma decay of the Br, Rb, Zn, Fe, Se and K in Blood of Colorectal Patients by INAA and PIXE appropriate radioisotopes was measured with a Ge(Li) detector. The irradiating and counting conditions are presented in Table 2. Known amounts of liquid and solid standards as well as blank samples (polyethylene irradiation containers containing 0.5 g anticlot) were irradiated together with the Fe monitors and under the same conditions. The p~ecison of the standards, comprising preparation and measurement, was checked for 3 different standards (one liquid and two solids), 4-6 samplings each, and found to be ± 2-6% for all elements studied. The detection limits (D.L.) defined as 3-JB, where B is the number of background counts in the region of interest, were calculated and found to be 0.061lg for Br and 0.09 Ilg for Rb. D.L.) The very low blank values found for Br and Rb could be considered as proof that sample preparation was performed in blank free conditions. «< P/XE The specimens for PIXE analysis were prepared by pipetting a certain volume of the blood or blood fraction (plasma or red blood cells) onto a pretreated 1.5 11m thick KIMFOL polycarbonate backing film, which had been mounted on a 2.5 cm diameter target ring. The target rings were pretreated so as to make the KIMFOL film hydrophillic. The procedure consisted of the following steps: first, 100 ilL concentrated suprapur HN03 acid was pi petted into the center of the film; after 24 hours or more, the HN0 3 was washed out with Milli-Q water, 40 ilL of a 0.05% solution ofpolyvinylpyrolidone in Milli-Q water was pipetted into the center, and the film was allowed to dry. For the red blood cells, 10 IlL'was then applied per specimen, 20 ilL in the case of the blood, and 40 ilL for the plasma. The pipetting of the blood and plasma was done using several volumes of 10 ilL, and the next 10 ilL was not added until the residue from the previous pipetting was nearly dry. The purpose of this sequential pipetting was to keep the diameter of the residue confined to less than 8 mm. The specimens were allowed to dry in the air, and, finally, they were coated with a 25 nm thick layer of carbon using a standard electron microprobe vac- Table 2. Characteristic data of the INAA working conditions Ele- Iso- E-keV Tin ment tope Irrad. time (h) Cooling time (d) Counting time (h) 82Br 554 86Rb 1077 42K 1530 59Fe 1099 65Zn 1115 75S e 136 3 14" - 35" 3 14" - 35 b 14" - 35" 14" - 35" 6 12' - 20b 6 12" - 20 b 12' - 20 b 12" - 20b 3 10' - 20 b 3 10' - 20b 10" - 20b 10" - 20" Br Rb K Fe Zn Se 35.34 h 18.70 d 12.36 h 45.10 d 244.00 d 120.00 d " Conditions for whole blood and red cells " Conditions for plasma 195 uum deposition apparatus. The carbon coating was applied to prevent charging up of the specimens during the PIXE bombardment. The various steps in the specimen preparation, with the exception of the carbon coating, were done on a clean bench. Three specimens were prepared from each individual san1ple and subjected to PIXE analysis. In addition to the actual sample specimens, blank specimens were also prepared and analyzed. The PIXE analyses of the specimens were done using an experimental setup and quantification procedures which have been described before (17-19). All bombardments were performed in vacuum using a 2.4 Me V proton beam of 8 mm diameter. A 660 11m thick Mylar absorber was interposed between the specimen and the Si(Li) detector. The beam current at the specimen was typically 150 nA, and the preset charge was 90 IlC per specimen. The PIXE spectra were analyzed for K, Fe, Cu, Zn, Se, Br and Rb. As no intemal standard had been added, concentration ratios to Zn were calculated, and these were then multiplied by the Zn concentration as obtained by INAA, in order to obtain the actual concentrations for the various elements. The detection limits of the PIXE analysis depended upon the element and the sample matrix, but were down to 0.3-0.4llg/g dry weight for the elements from Fe to Br. Results and Discussion This work summarizes the results of a preliminary study regarding the concentration of Br, Rb and some other trace elements in blood and blood fractions of CRC patients. Two nondestructive, complementary methods were applied - INAA and PIXE. The possibility of using the PIXE method instead of INAA was checked. Table 3 Table 3. INAAjPIXE concentration ratios" in blood (B), red cells (R) and plasma (P) samples from normals (N) and cancer patients (C) Br N C Rb N C Br/Rb N t K Fe Se N C N C N C B R P 0.936 ± 0.195 0.935 ± 0.106 0.796±0.103 0.807 ± 0.161 1.201 ± 0.191 1.165 ± 0.214 1.042 ± 0.183 0.875 ± 0.180 1.270 ± 0.090 1.212 ± 0.105 1.261 ±0.190 1.156 ± 0.096 1.072 ± 0.197 0.974 ± 0.164 0.894 ± 0.080 0.970 ± 0.060 1.164 ± 0.247 0.993± 0.186 1.113 ± 0.073 0.973 ± 0.090 1.192 ± 0.055 1.181 ±0.100 1.382 ± 0.072 1.024 ± 0.059 0.941 ± 0.152 0.940 ± 0.167 1.017 ± 0.351 0.895 ± 0.351 " Mean INAAjPIXE ± S.D. of individual results 0.976 ±0.139 1.171 ± 0.180 1.058 ± 0.134 196 C. Shenberg, H. Feldstein, R. Comelis, L. Mees, J. Versieck, L.VanbaUenberghe, J. Cafmeyer and W. Maenhaut Table 4. Elemental concentration ratios between blood (B), red cells (R) and plasma (P) in normals and cancer cases obtained by INAA and PIXE analysis! Element Br Rb Fe Se K Zn Cu INAA PIXE INAA PIXE INAA PIXE INAA PIXE INAA PIXE INAA PIXE B/R-N B/R - C B/P-N B/P - C 1.275 ± 0.008 1.482 ± 0.196 0.610 ± 0.049 0.675 ± 0.048 0.575 ± 0.064 0.542 ± 0.055 0.939 ± 0.084 1.024 ± 0.151 0.595 ± 0.082 0.673 ± 0.074 0.618 ± 0.059 1.646 ± 0.331 1.374±0.124 1.445 ± 0.101 0.535 ± 0.066 0.628 ± 0.115 0.472 ± 0.098 0.459 ± 0.095 0.830 ± 0.095 0.788±0.121 0.542 ± 0.098 0.624 ± 0.110 0.540 ± 0.079 1.628 ± 0.404 0.843 ± 0.139 0.777 ± 0.084 9.215 ± 2.930 10.616 ± 3.323 337.57 ± 66.50 354.92 ± 94.65 1.403 ± 0.063 l.240± 0.148 8.702 ± 0.792 8.455 ± 0.482 7.515 ± 0.541 1.067 ± 0.154 0.780 ± 0.069 0.755 ± 0.185 10.624 ± 3.624 11.179 ± 4.896 487.78 ± 139.99 467.47 ± 95.36 1.320± 0.183 1.175 ± 0.123 7.773 ± 1.819 10.118 ± 3.349 7.885 ± 1.001 0.918 ± 0.159 [ Mean ± S.D. of individual B/R and B/P ratios, representing all cases studied; N = normals, C = cancer cases presents the INAA/PIXE concentration ratios obtained for the element studied. As can be seen, fairly good agreement between the two methods was obtained for plasma samples (± 10%), while the differences for Rb, Se and Fe in whole blood and red cells were ± 20%. The reasons for this discrepancy are not fully clear, but part of it may be due to a systematic error in the correction for matrix effects in PIXE, and for PIXE, Rb spectral interferences may also have contributed. Nevertheless, as the differences between the two methods are systematic, in the comparison of trace elements in the blood of healthy people and CRC patients, which was the main aim of the present study, similar results were obtained by both methods. Our previous study (16) was conducted on whole blood samples only. Therefore, it was of interest to study the distribution of trace elements in the blood fractions, red cells and plasma. Table 4 presents the concentration ratios of the elements among whole blood and blood fractions, plasma and red cells, measured by both methods. No significant differences were observed in the above ratios in either of the groups studied : normals and cancer cases. Very close results were also obtained with both methods of determination: INAA and PIXE. Table 5. Accuracy (A) of the INAA method. Analysis of NBS Bovine Liver 1577a and Certified Freeze-Dried Human Serum The precision of the INAA method calculated from duplicates of 69 samples for Br determination (all fractions included), 46 whole blood and red cells samples for Rb determination and 23 plasma samples for Rb, was ± 3.2%, ± 3.1 %, ± 5.3%, respectively. To check the accuracy of the INAA technique, Br and Rb liquid standards were prepared and analyzed and further compared with solid serum and bovine liver certified standards. For Se and Zn the biological reference material (freeze-dried human serum) served as a standard for calculating the concentration of these elements in bovine liver and the results were compared. The results with the certified values reported by the National Bureau of Standards are given in Table 5. As can be seen, a good agreement was found for the Br and Rb content in serum and bovine liver and for Zn in bovine liver. Se analyzed in the same way as Zn differed from the reference value, but Table 6. Representative results (mg/L) of bromine analyses in whole blood samples a ; comparison between measured values and values calculated from the hematocrit content b Br in mg/L Sample Hemat INAA % Meas. Calc. I Element Certified amount mg/kg Found mg/kg Bra Brb Rb o Rb b Se b Zn b 48.48 ± 1.55 9 1.67 ± 0.11 12.5 ± 0.1 0.71 ± (J.07 123 ±8 46.44 ± 9.05 ± 1.564 ± 13.566 ± 0.833 ± 134.79 ± A% 1.95 0.55 0.034 0.977 V.087 10.51 - 4.2 + 0.6 - 6.3 + 8.5 +17.3 + 9.6 Biological Reference Material (freeze-dried human serum) (24) Standard Reference Material 1577a (Bovine Liver) National Bureau of Standards, Washington D.C. 20234, U.S.A. 2 3 4 5 '6 7 8 9 10 41.6 37.3 35.9 43.0 43.1 41.3 32.6 43.1 27.7 47.7 4.162 2.430 2.880 3.239 3.130 3.000 1.470 2.780 1.835 2.338 3.815 2.692 2.790 2.890 3.115 3.007 1.438 2.708 2.021 2.393 PIXE ~% + 8.3 - 10.8 + 3.1 +10.8 + 0.5 - 0.2 + 6.1 + 2.6 - 10.1 - 2.4 ~% Meas. Calc. 2.940 2,880 2.700 3.061 2.901 2.489 - 4.1 - 0.7 + 7.8 2.904 1.656 2.732 1.739 2.780 2.971 1.615 2.851 1.823 2.829 - 2.3 +2.5 - 4.4 - 4.8 - 1.8 Samples 1-4 normals; 5-10 cancer cases Results provided by the Laboratory for Clinical Biology, University Hospital, Gent a a b b Br, Rb, Zn, Fe, Se and K in Blood of Colorectal Patients by INAA and PIXE Table 7. Representative results (mg/L) of potassium in human semm; comparison between PIXE and ISE (Ion Specific Electrode)' Table 8. Trace element determination (mg/L) in the plasma of the normal population in Belgium Mean ± S.D. K(mg/L) Sample N° ISE PIXE !J.% 1 2 3 4 5 6 7 8 9 10 11 12 13 156.8 165.4 144.3 156.8 152.1 142.3 141.2 167.7 147.4 172.0 171.2 134.5 180.6 157.8 159.0 165.2 165.3 167.3 126.6 131.4 141.6 146.1 176.7 165.3 166.4 183.2 + 0.6 - 4.0 + 12.6 + 5.1 + 9.1 - 12.4 - 7.4 - 18.4 - 0.9 + 2.7 - 3.6 + 19.2 + 1.4 , Results provided by the Laboratory for Clinical Biology, University Hospital, Gent that may be explained by low counting statistics and the relatively high errors given for the certified standard. The accuracy of INAA and PIXE methods was further tested. Based on the hematocrit content measured by the Coulter technique with a Coulter STKS (Coulter Electronic), we calculated the Br concentration in whole blood from its concentratIon in plasma and red cells and compared it with values measured by INAA and PIXE. The results are presented in Table 6. The agreement between measured and calculated values indicates the reliability of both methods used for Br determination. Table 7 presents the results of K in plasma, obtained by PIXE method and compared to K values measured with ISE, an Ion Specific Electrode (Hitachi 747 Boehringer, Manheim). Except for samples 8 and 12, all other results are in quite good agreement for both methods. Br, Rb, Se and Zn levels in the plasma of normal populations in Belgium were reported by several laboratories and the results, together with ours, are shown in Table 8. As can be seen, the PIXE and INAA results obtained in the present work are in a good agreement with others. The statistical analysis (Student t-test) of the final results of elemental concentrations obtained for whole blood samples from normals and cancer cases with both methods, is given in Table 9. For the 11 cancer cases investigated, the mean values of Br, Rb, BrlRb, Fe, K and Se were significantly lower than for the healthy individuals. (p<O.Ol, <0.001, <0.05, <0.01, <0.05, <0.05, respectively). These results are in agreement with the findings of our previous study (16);, with one exception: the BrlRb for the Israeli population was much higher than for the Belgian. The reason for this deviation could be due to the large difference in Br levels in the blood of the two populations: approximately 4.5 mglL in Belgium and 8-10 197 Range Bf" 4.870 ± 2.020 1.280 - 7.480 4.090 - 4.780 4.440 5.770 ± 0.550 5.360 W of cases Analytical methods 10 poolof27 27 100 INAA INAA and PIXE PIXE PIXE Bf" 4.309 ± 1.379 3.080 -7.576 10 4.398 ± 1.306 3.260 - 6.291 10 INAA PIXE Rb" 0.170 ± 0.040 0.090 - 0.270 10 INAA 0.168 0.138 - 0.198 poolof27 AAS,AES,INAA Rbb 0.212 ± 0.053 0.148 - 0.305 10 0.249 ± 0.109 0.129 - 0.433 10 INAA PIXE se 0.130 ± 0.020 0.090 - 0.180 0.073 ± 0.012 0.097 ± 0.012 0.067 - 0.123 0.110 ± 0.032 0.095 0.091 - 0.100 36 40 163 27 pool of 27 0.097 ± 0.014 0.084 ± 0.012 74 145 INAA HAAS HAAS prXE AAS,FLU,IDMS, INAA,PIXE GFAAS GFAAS Sen 0.087 ± 0.013 0.071 ± 0.014 10 10 INAA PIXE Zn' 0.940 ± 0.130 0.690 - 1.210 0.850 ± 0.11 0 1.100 0.854 - 0.891 0.873 46 27 100 pool of 27 RNAA PIXE PIXE AAS,ICP-AES, INAA,PIXE,VTM AAS 0.922 ±0.137 28 Znb 0.750 ±O.llO 0.580 - 0.926 10 INAA Cu' 1.070 ± 0.240 0.730 - 1.990 46 27 0.950 ± 0.380 100 1.100 0.973 - 1.045 poolof27 1.009 RNAA prXE PIXE AAS,ICP-AES, IN AA,PIXE,VTM AAS 1.170 ± 0.160 28 Cu" 1.095 ± 0.391 0.775 - 1.846 8 " See ref. (21), b PIXE Present work mglL in Israel. No difference was found for the Zn content in samples from normal and cancer cases and the Cu concentration was found to be slightly higher in cancer cases (p<0.3), which is in good agreement with the results reported in the literature (8). Conclusions Interesting preliminary results were obtained concerning the levels of trace elements in the blood of CRC patients. The concentrations of Br, Rb, Fe, K and Se were significantly lower relative to nornlals with p values ranging from 0.05 to 0.001. Patients 1 and 3 were sam- 198 C. Shenberg, H. Feldstein, R. Comelis, L. Mees, 1. Versieck, L.VanbaUenberghe, J. Cafmeyer and W. Maenhaut Table 9. INAA and PIXE results of elemental concentrations (mg/L) in the whole blood samples of normal (n = 10) and cancer cases (n 13) , = Concentration mglL INAA Br Rb Br/Rb Se Zn Cu Fe K PIXE Nomlals Cancer NOimals Cancer pb 3.863 ± 1.205 1.882 ± 0.235 2.085 ± 0.709 0.123 ± 0.019 5.627 ± 0.777 2.427 ± 0.748 1.432 ± 0.285 1.557 ± 0.413 0.100 ± 0.025 5.752 ± 0.949 3.379 ± 0.916 2.354 ± 0.363 1.842 ± 0.803 0.086 ± 0.020 2.508 ± 0.742 1.775 ± 0.540 1.392 ± 0.362 0.078 ± 0.026 < < < < 1.053 ± 0.243 1.227 ± 0.218 <0.3 425.4 ± 57.8 1563.5±267.4 < 0.05 < 0.01 582.5 ± 43.7 1496.0± 145.1 487.7 ± 112.7 1275.3 ± 181.2 461.3 ± 55.4 1413.3 ± 6.9 0.01 0.001 0.05 0.05 " Mean values ± S.D.; each individual determination is the mean of two samplings. b Variance according to Student's t-test based on the INA A results obtained for normals and cancer cases, except for Cu which was based on PIXE results. pled twice with an interval of a few weeks between each sampling. No significant changes in the concentration of the elements studies were observed . . As our earlier work dealt mainly with the BrlRb ratio in the blood of eRe patients in Israel (16), we focused our interest on those two elements in the present study as well. Whereas for 11 cancer cases investigated, a mean value of 2.43±O.78 rrig/L was assessed for Br and 1.43±O.28 mg/L for Rb, the values for 10 controls covered a range of 3.86±1.20 mg/L and 1.88±O.23 mg/L, respectively. The BrlRb ratio found was higher for the control group, as compared to the patients (p<O.05), which corroborates our expectations based on the previous work (16). The only exception was the much lower BrlRb ratio in Belgium than in Israel. Such results could be explained by the influence of dietary habits and environmental factors on the Br level in the blood of both populations (21); the Br content in the blood of the Israeli population is twice as high as in that of the Belgian citizens (22,23). For some elements in whole blood and red cells there was a systematic difference between PIXE and INAA of ±20%, but otherwise the two methods gave an agreement to within 10%. For further studies we recommend the use of a combined PIXE - INAA method. INAA will determine quantitatively Bf and K, based on relatively short irradiation, cooling and counting times ( approximately 9 h, 5 days, 9 h, respectively). The same samples will be analyzed for Rb and a few other elements by PIXE, and the absolute values calculated according to the results obtained by the INAA method. The ratios of the concentrations in whole blood to red cells and whole blood to plasma were not significantly different for normals and cancer cases and therefore in future studies the analysis of whole blood only may be sufficient. The significance of the difference in the bromine and rubidium content in eRe patients has not yet been determined. Biological interest in rubidium has been stimulated by its close physicochemical relationship to potassium. Although the role of Br and Rb in colorectal cancer is unclear, the present results may be of importance as an additional parameter for differentiating between normal and malignant cases. Further experiments have to be undertaken in order to draw significant conclusions. References 1. SCHRAUZER, G.N. (1984) The discovery of the essential trace elements. An outline of the history of biological trace element research. In: Biochemistry of the essential ultratrace elements. Plenum Publishing Co. , New York pp. 17-31 2. BRATTER, P., NEGRETTI DE BRATTER, VE. ROSICK, U. and V STOCK-HAUSEN, H.B. (1987) Trace element concentration in serum of infants in relation to dietary sources. In: Trace Element Analytical Chemistry in Medicine and Biology, Vol. 4 (Brauer, P. and Schramel, P. eds.) Walter. de Gruyter, Berlin, pp. 133-143 3. PARR. R.M. (1987) An international collaborative research program on minor and trace elements in total diets. In: Trace Element Analytical Chemistry in Medicine and Biology, Vol. 4 (Bratter, P. and Schramel, P. eds.) Walter de Gruyter, Berlin, pp. 157- 160 4. SCHRAUZER, G.N. (1987) Trace elements in cancer diagnosis and therapy. A review. In: Trace Element Analytical Chemistry in Medicine and Biology, Vol. 4 (Brauer, P. and Schramel, P. eds.) Walter de Gruyter, Berlin, pp. 403-417 5. NELSON, R.L. (1990) Dietary minerals and colorectal cancer; a review. In: Metal Ions in Biology and Medicine (Callery, Ph. Poirier, L.A. Manfait,M . and Etienne J.e. eds.) John Libbey Eurotext, Paris, pp. 35-39 6. RIZK, S.L. and SKY-PECK, H.H. (1984) Comparison between concentrations of trace elements in normal and neoplastic human breast tissue, Cancer Res. Vol. 44, pp. 5390-5394 Br, Rb, Zn, Fe, Se and K in Blood of Colorectal Patients by INAA and PIXE 199 7. KWAN-HOONG, N.G., BRADLEY, D.A., LAI-MENG LOO!', SEMAN MAHMOOD, C. and KHALIK WOOD, A. (1993) Differentiation of elemental composition of normal and malignant breast tissue by instrumental neutron activation analysis, App!. Radiat. Isot. Vo!. 44, pp. 511-516 16. SHENBERG, C., SPIEGEL, S., CHAITCHIK, S., JORDAN, P., KITZIS, M., ALTMAN, M., COHEN, J. and BOAZI, M. (1994) The distribution of BrKa/RbKa ratio in human whole blood; comparison between normal and colorectal cancer cases. BioI. Trace Elem. Res. 42, pp.231-241 8. DUROSINMI, M.A., 010, J.O., OLUWOLE, A.F, ONONYE, A.F, AKANLE, O.A. and SPYROU, N.M. (1994) Study of trace elements in blood of cancer patients by proton-induced Xray emission (PIXE) analysis, Bio!. Trace Elem. Res. 43-5, pp. 351-355 17. MAENHAUT, w., DE REU, L., VAN RINSVELT, H.A., CAFMEYER, J. and VAN ESPEN, P. (1980) Pruticle induced X-ray emission (PIXE) analysis of biological materials: precision, accuracy and application to cancer tissues. Nuc!. Instr. Meth, 168, pp. 557-562 9. CORNELIS, R., RINGOIR, S., LAMElRE, N., MEES, L. and HOSTE, J. (1979) Blood bromine in uremic patients. In: MineraI Elecu'olyte Metab., Vo!. 2, Massry,S.G. ed, Karger, S. Basel, pp. 186-192 18. MAENHAUT, W. and RAEMDONCK, H. (1984) Accurate calibration of a Si(Li) detector for PIXE analysis. Nuc!. Instr. Meth. Phys. Res., Bl, pp. 123-136 10. PETRINI, M., VAGLINI, Y. and CARULLI, G. (1986) Effect of lithium and rubidium on the differentiation of mononuclear cells, Int. J. Tissue React, 8, pp. 391-392 11. MALEK, A.P. and WILLIAMS, J.A. (1984) Rubidium in psychiatry : research implications, Pharmacol, Biochem. Behav, 21, Supp!. I. pp. 49-50 12. PLACID!, G., LENZI, A., LAZZERINI, F (1988) Exploration of the clinical profile of rubidium chloride in depression - a systematic open trai!. J. Clin. Psychopharmaco!. 8, pp. 184-188 13. PFLUG, B., KOHLER, W. and CARELLA, A.B. (1990) Effect of rubidium on the course of illness and the circadian rectal temperature in a 66 year old, depressive woman. Chronobio!. Int, 7, pp: 463-465 19. MAENHAUT, w., VANDENHAUTE, J. and DUFLOU, H. (1987) Applicability of PIXE to the analysis of biological materials, Fresenius Z. Ana!. Chem. 326, pp. 736-738 20. HUTTER, R.Y. and SOBIN, L.A. (1986) A universal staging system for cancer of the colon and rectum. Let there be light. Arch. Patho!' Lab. Med. 110, pp. 367-368 21. CORNELIS, R., SABBIONI, E., VAN der VENNE, M.T. (1994) Trace elements reference values in tissues from inhabitants of the European Community. Vll. Review of trace elements in blood serum and urine of the Belgian popUlation and critical evaluation of their possible use as reference values. The Science of the Total Environment, 158, pp.191-226 22. VERSIECK, J. and CORNELIS, R. Normal levels of trace elements in human blood plasma or serum - Review (1980) Ana!. Chim. Acta, 116, pp. 217-254 14. SHENBERG, c., MANTEL, M., GILAT, J., CHAITCHIK, S., STADLER, J. and ALON, R. (1989) Comparison between trace elements in normal and tumorous human colon tissue. In: Israel Atomic Energy Commission Annual Report, IA-I448, pp. 145146 23. RAPAPORT, M.S., MANTEL, M. and SHENBERG, C. (1982) Determination of bromine in blood serum by 125 1 excited X-ray fluorescence. Med. Phys. 9(2), March-April, pp.194-198 15. SHENBERG, c., BOAZI, M., COHEN, 1., KLEIN, A., KOJLER, M., NYSKA, A. and RAVE, S. (1994) BrlRb ratio obtained by XRF analysis of kidneys of normal and tumour bearing mice treated with cis-DDP. J. Trace Elem. Health Dis. 8, pp.I77-182 24. VERSIECK, J., VAN BALLENBERGHE, L., DE KESEL, A., HOSTE, J., WALLAEYS, B., VANDENHAUTE, J., BAECK, N., STEYAERT, H., BYRNE, R.A., SUNDERMAN, FW., Jr. (1988) Certification of a second-generation biological reference material (freeze-dried human serum) for trace element determinations. Ana!. Chim. Acta, 204, pp. 63-75