- Email: [email protected]
Chromosome aberrations induced in human lymphocytes by in vitro and in vivo X-rays
Mutation Research 517 (2002) 167–172
Chromosome aberrations induced in human lymphocytes by in vitro and in vivo X-rays Heike Schröder a,b,∗ , Anna Heimers a,b a
b
Center of Environmental Research and Technology (UFT), University of Bremen, Leobener Strasse, 28359 Bremen, Germany Bremen Institute for Prevention Research, Social Medicine, and Epidemiology (BIPS), Linzer Strasse 8, 28359 Bremen, Germany Received 16 October 2001; received in revised form 14 March 2002; accepted 18 March 2002
Abstract A dose–effect curve is presented obtained by analysis of dicentric chromosomes and centric ring chromosomes in lymphocyte metaphase spreads of three healthy volunteers after in vitro 100 kV X-ray-irradiation of peripheral blood samples. This calibration curve follows a linear quadratic equation, y = c + αD + βD 2 , with the coefficients: y = (0.0005 ± 0.0001) + (0.0355 ± 0.0066)D + (0.0701 ± 0.0072)D 2 . The model is based on 13.231 first-division metaphases analyzed after in vitro exposure to doses ranging from 0.1 to 2.0 Gy at a dose rate of 0.4 Gy min−1 . Significant overdispersion of the observed chromosomal aberrations was evident for dose points 1.0 and 2.0 Gy, respectively. The calibration curve was applied to derive equivalent whole body doses of three subjects after suspected extensive exposure to diagnostic X-rays. © 2002 Elsevier Science B.V. All rights reserved. Keywords: 100 kV X-rays; Dose–effect curve; Human lymphocytes; Chromosome aberrations; Dicentric chromosomes; Biological dosimetry
1. Introduction The analysis of dicentric chromosomes and centric ring chromosomes is considered the most sensitive biological method to quantify exposure to radiation. In the absence of physical measurements biological dosimetry is the method of choice to prove or exclude whether an individual or a group had been affected by ionizing radiation in the past. To accurately quantify the individually absorbed dose it is a prerequisite to establish a representative calibration curve in a standardized in vitro experiment [1]. In medical diagnostics X-rays are applied most frequently with energies of about 100 kV. Therefore, we ∗
Corresponding author. Tel.: +49-421-218-7617; fax: +49-421-218-7616. E-mail address: [email protected] (H. Schröder).
decided to perform this calibration experiment with 100 kV X-rays in such a way that it should enable us to present a dose–effect curve applicable to diagnostic radiography. Chromosome aberration analysis has also been done in three persons who were suspected of an extensive previous exposure to diagnostic X-rays. Based on the calibration curve their equivalent whole body doses were derived from the aberration frequencies observed.
2. Materials and method 2.1. Irradiation conditions Peripheral blood samples were taken by venipuncture from three healthy volunteers in the radiological practice of Dr. Gerhard Schneider and co-workers,
1383-5718/02/$ – see front matter © 2002 Elsevier Science B.V. All rights reserved. PII: S 1 3 8 3 - 5 7 1 8 ( 0 2 ) 0 0 0 6 7 - 0
168
H. Schröder, A. Heimers / Mutation Research 517 (2002) 167–172
Wuppertal, Germany, where the 100 kV X-ray source, RT 100 Müller, (0.4 Gy min−1 , 8 mA, 2.0 mm Al) was located. If not involved in the irradiation procedure, blood samples were kept at 37 ◦ C in a waterbath all the time prior to culture. The irradiation of the samples took place after exact positioning of the probes in an acrylic–glass phantom, which simulated a tissue depth of 4.5 cm. Doses of 0.1, 0.2, 0.5, 1.0 and 2.0 Gy were applied to the whole blood samples. 2.2. Culture conditions Due to pragmatic requirements (irradiation could just be performed after the routine work of the practice had finished and the samples had to be transferred from Wuppertal to Bremen) 18 h after irradiation lymphocyte cultures were set up in the Bremen laboratory according to a standardized cell cycle controlling method described elsewhere [2]. This includes the use of Bromodeoxyuridin which enables the analysis of exclusively first division metaphases. Harvesting of lymphocyte metaphases followed after 48 h culture time. 2.3. Cytogenetic analysis After Fluorescence plus Giemsa staining, chromosome aberration analysis was applied to only first division metaphases. Collection of metaphases was facilitated by an automatic computerized system including a tool of data management (MetaSystems, Altlussheim, Germany). According to our laboratory standards only complete metaphases with 46 centromers were included in the analysis. To analyze exchanges/translocations all chromosomes were assigned to their respective groups according to the International System for Human Cytogenetic Nomenclature [3]. All metaphases were documented and those with suspected chromosomal aberrations were karyotyped routinely for verification. 2.4. Investigated subjects Female, 50 years old, suspected an overdose of X-rays after three diagnostic thorax radiograms within 3 months. After the last one she went under medical care and a radiation dermatitis was supposed. To eval-
uate this diagnosis a blood sample was taken for biological dosimetry 3 months after the last radiogram. Female, 40 years old, also suspected an overdose of X-rays after one diagnostic thorax radiogram because after that she felt ill with various unspecific complaints and a leucopenia had been diagnosed. An independent expert laboratory in the Netherlands calculated a dose of about 600–6000 mGy. 2.5 years later a blood sample has been taken for biological dosimetry to evaluate the radiation dose. Male, 54 years old, had a diagnostic radiation exposure of about 88 mSv, which was documented and had been received during the last 4 years before blood sampling for biological dosimetry. Since this man was in fear of having received an overdose chromosome aberration analysis was performed in order to evaluate the radiation exposure. Dicentric frequencies were compared with the Bremen laboratory control of 0.0005 (95% CI: 0.0003– 0.0008). The equivalent whole body doses were computed based on the in vitro X-ray dose–effect curve. 2.5. Statistics The parameters for the dose–effect curve were obtained by least squares regression, using iteratively reweighted inverse Poisson variances as weights. Goodness of fit was tested with the Pearson χ 2 -value. The confidence intervals were calculated using the method of [4].The variance by mean ratio was used to test for Poisson distribution. Values of (variance/ mean) < 1 indicate underdispersion while values >1 indicate overdispersion. Overdispersion is significant (P < 0.05) when the test quantity u exceeds 1.96 [5,6]. Fisher’s exact test was applied for analyzing differences in dicentric frequencies between investigated subjects and the laboratory control.
3. Results In a total of 13.231 analyzed metaphase spreads 424 dicentric chromosomes, six tricentric chromosomes (tric) and 28 centric ring chromosomes (cR) were found. Counting one tric as two dicentric equivalents, a total of 465 of these unstable chromosome type rearrangements was observed. The percentage
H. Schröder, A. Heimers / Mutation Research 517 (2002) 167–172
169
Table 1 Cytogenetic results of the chromosome aberration analysis obtained after 100 kV X-ray irradiation Dose (Gy)
No. of cells analyzed
No. of dic
No. of cR
Mean dic + cR
95% CI−
95% CI+
No. of excess ace + min
0 0.1 0.2 0.5 1.0 2.0
3013 3014 3020 2063 1333 788
5 22 25 68 110 207
0 0 0 4 7 17
0.0017 0.0073 0.0083 0.0349 0.0878 0.2843
0.00054 0.00457 0.00536 0.02731 0.07259 0.24825
0.00387 0.01105 0.01222 0.04395 0.10519 0.32403
14 49 63 144 227 400
dic: dicentric chromosome; cR: centric ring chromosome; 95% CI−: lower 95% confidence interval (according to the Poisson distribution); 95% CI+: upper confidence interval (according to the Poisson distribution); ace + min: excess acentric fragments and double minutes.
of cR of the total is 6.2%. Table 1 gives the results for the different dose points. Dicentrics and centric rings have been fitted to the linear–quadratic dose effect model: y = c + αD + βD 2 , and it yielded the following coefficients: y = (0.0005 ± 0.0001) + (0.0355 ± 0.0066)D(0.0701 ± 0.0072)D 2 (χ 2 = 6.9; 3 d.f.). Fig. 1 shows the appropriate dose–effect curve with the 95% confidence intervals.The intercellular distribution of the observed dicentric chromosome equivalents and centric ring chromosomes is given in Table 2. The dicentric data show overdispersion after 1.0 and 2.0 Gy. Table 3 presents the results of the chromosome aberration analyses of the three individuals and gives the corresponding doses. Each subject shows a significant increase (P < 0.05) of dicentric chromosomes compared to our laboratory control.
Fig. 1. Dose–effect curve for 100 kV X-rays. The fitted values of the coefficients of the linear quadratic function y = c + αD + βD 2 are: c = (0.05 ± 0.01) × 10−2 ; α = (3.55 ± 0.66) × 10−2 ; β = (7.01 ± 0.72) × 10−2 . The 95% confidence intervals are represented by dashes.
Table 2 Intercellular distribution of dicentric chromosomes and centric ring chromosomes after 100 kV X-ray irradiation Dose (Gy)
Distribution 0
1
0 0.1 0.2 0.5 1.0 2.0
3008 2992 2995 1993 1226 606
5 22 25 68 97 149
∗
2
2 10 25
3
7
4
1
Variance/ mean 0.999 0.994 0.992 1.021 1.084 1.182
u
– – 0.68 2.18∗ 3.61∗
Overdispersion is significant, P ≤ 0.05.
4. Discussion In vitro calibration curves are prerequisites to quantify absorbed doses in exposed individuals or groups. Different dose–effect curves after low LET radiation show that the intercellular distribution of dicentric chromosomes follows a Poisson distribution [7–10]. However, the dose–response curve for 100 keV X-rays presented here shows a significant overdispersion in two of five doses in fact after 1 and 2 Gy. The coefficients agree well with other laboratories [8–11]. There is no full explanation of the observed overdispersion. Unlike other working groups in this experiment overdispersion was observed at the highest doses. Even if tricentric chromosomes and centric ring chromosomes are excluded from the analysis (like other laboratories do) overdispersion still remains statistically significant for both dose points. As the irradiation conditions in our experiment, i.e. dose, dose-rate, radiation quality, and temperature are comparable to others, these factors may not attribute to the observed
170
H. Schröder, A. Heimers / Mutation Research 517 (2002) 167–172
Table 3 Investigated subjects Subject
Age (years)
Time of blood sampling
No. of analyzed cells
No. of dicentrics, mean/cell (95% CI)
Equivalent whole body dose, Gy (CI)
Female
50
3 months after irradiation
1028
3 0.003 (0.0006–0.0085)
0.063 (0.002–0.195)
Female
55
2.5 years after irradiation
1008
5a 0.005 (0.0016–0.0116)
0.105 (0.024–0.250)
Male
54
Immediately after 4 years of diagnostic X-ray radiation exposure
802
4
0.105
0.005 (0.0016–0.0116)
(0.024–0.250)
a
One cell with two dicentrics.
phenomenon. A difference to other published curves is the fact that our cultures had been established around 18 h after the irradiation procedure. But according to Darroudi et al. [12] dicentric chromosomes are formed immediately after irradiation and do not change with time after X-ray doses up to 2.0 Gy. Therefore, the prolonged time between irradiation exposure and culture may be excluded as the cause of the observed deviation from the Poisson distribution. However, this experiment was conducted using the tube voltage of 100 kV instead of the most commonly used 220 kV, and higher effects like increased values of the relative biological effectiveness (RBE) [13,14] were observed after irradiation with soft X-rays. We suppose that the applied 100 kV X-ray irradiation might be a possible reason for the observed overdispersion. By reanalysis the data of the cobalt-60 gamma dose– response curve obtained in our lab with the iteratively reweighted least squares method the linear–quadratic equation has the following coefficients: y = 0.0004 ± 0.0001+0.0064±0.0040D +0.0480±0.0048D 2 . According to Schmid and Bauchinger [15] this method is reliable for evaluating biological effects after low-LET radiation. The ratio of the α-coefficients gives an RBE for low dose X-rays of 5.55±3.62 compared to Co-60 gamma which is significant at the 5% level. Hence the linear term increases with decreasing photon energy. These data show an effect which is discussed at present in terms of radiation protection [16]: the different biological effectiveness of various low-LET radiations. The evaluation of experimental data on the
induction of dicentric chromosomes by photons and electrons yielded a factor of about 6 decrease in effectiveness between Cr X-rays and 3 MeV electrons [15]. According to Straume [17], the use of the single risk and WR or Q-value 1 for all low LET radiations may be inappropriate. In our experiment, acentrics show overdispersion in four of five doses (data not shown). Lloyd et al. [10] also found overdispersion in acentric data after X-ray irradiation. Schmid et al. [8] observed overdispersion for acentrics in five of nine doses in one experiment after X-ray irradiation but Poisson distribution in another experiment. The difference was due to variances in the filter combinations. Schmid et al. [8] pointed at the influence of the X-ray spectrum due to filter conditions applied and concluded that the differences in the dose–effect curves appear to be mainly a consequence of different X-ray bremsstrahlung spectra. This may also be an explanation for the observed overdispersion of dicentric chromosomes after 1 and 2 Gy. Studies on the induction of chromosome aberrations by diagnostic X-rays have been done by Weber et al. [18]. They determined dicentric yields per cell in the range of 0.0037–0.022 (mean: 0.0095) for five patients after extensive exposure. This was an unexpectedly high yield of dicentrics but the authors explained most of the findings with the additional use of iodized radioactive contrast media and the application of computed tomography. All investigated subjects in the present study show a significant elevation of dicentrics compared to the
H. Schröder, A. Heimers / Mutation Research 517 (2002) 167–172
Bremen laboratory control. The dicentric yield of 0.003 and the mean equivalent whole body dose of 63 mGy in subject 1 confirmed the suspicion about a radiation dermatitis. Also the suspected overdose of X-rays allegedly after only one diagnostic thorax radiogram could be verified by biological dosimetry. A dicentric yield of 0.005 in subject 2 leads to a mean equivalent whole body dose of 105 mGy. The finding of one cell with two dicentrics shows the influence of partial body exposure. The dose range of 24–250 mGy has to be regarded as a minimal range because the halflife of lymphocytes and the decline of dicentric chromosomes with time has not been considered. A calculated dose of at least 600 mGy by an independent expert laboratory therefore is not unlikely. Subject 3 is of special interest because his X-ray exposure was well documented over the last 4 years before blood sampling and amounted a value of approximately 88 mGy. The observed dicentric yield of 0.005 gives a mean equivalent whole body dose of 105 mGy which agrees well with the computation based on the X-ray documentation. The dicentric frequencies in the first two cases show that both subjects had overexposures of diagnostic X-rays which were beforehand confirmed by medical reports. These two cases show that minimum requirements in radiation protection had not been fulfilled. An incorrect distance from the X-ray machine to the patient for example can lead to 10 times higher patient doses [19]. Diagnostic radiography requires safe and modern equipment and competent personnel as well as an unequivocal medical indication. Subject 3 has no overexposure from X-rays, the equivalent whole body dose, derived from the dicentric frequency, corresponds with the applied state of the art X-ray doses. In this case the different X-ray exposures given over 4 years could be detected by means of biological dosimetry, indicating that this is still the method of choice also in the low dose range.
Acknowledgements The authors are grateful to Gerhard Schneider and his staff for the opportunity to use their X-ray machine in order to conduct the in vitro irradiation series. They thank Herbert Braselmann and Achim Kranefeld
171
for statistical advice, Ingeborg Grell-Büchtmann, Bettina Dannheim and Boris Oberheitmann for their laboratory work, and Heiner v. Boetticher, Wolfgang Hoffmann and Inge Schmitz-Feuerhake for valuable discussion. References [1] M. Bauchinger, E. Schmid, S. Streng, J. Dresp, Quantitative analysis of the chromosome damage at first division of human lymphocytes after 60 Co ␥-irradiation, Radiat. Environ. Biophys. 22 (1983) 225–229. [2] A. Heimers, H. Schröder, E. Lengfelder, I. Schmitz-Feuerhake, Chromosome aberration analysis in aircrew members, Radiat. Prot. Dosim. 60 (1995) 171–175. [3] ISCN, In: F. Mitelman (Ed.), An International System for Human Cytogenetic Nomenclature, S. Karger, Basel, 1995. [4] C.J. Clopper, E. Pearson, The use of confidence or fiducial limits illustrated in the case of binomial, Biometrica 26 (1934) 404–413. [5] R.C. Radhakrishna, J. Chakravarti, Some small sample tests of significance for a Poisson distribution, Biometrics 12 (1956) 264–282. [6] J.R.K. Savage, Sites of radiation induced chromosome exchanges, Curr. Top. Radiat. Res. 6 (1970) 129–194. [7] A.A. Edwards, D.C. Lloyd, R.J. Purrott, Radiation induced chromosome aberration and the Poisson distribution, Radiat. Environ. Biophys. 16 (1979) 89–100. [8] E. Schmid, M. Bauchinger, S. Streng, U. Nahrstedt, The effect of 220 kVp X-rays with different spectra on the dose response of chromosome aberrations in human lymphocytes, Radiat. Environ. Biophys. 23 (1984) 305–309. [9] L. Fabry, A. Leonard, A. Wanbersie, Induction of chromosome aberrations in G0 human lymphocytes by low doses of ionizing radiations of different quality, Radiat. Res. 103 (1985) 122–134. [10] D.C. Lloyd, A.A. Edwards, J.S. Prosser, Chromosome aberrations induced in human lymphocytes by in vitro acute Xand gamma radiation, Radiat. Prot. Dosim. 15 (1986) 83–88. [11] J.F. Barquinero, S. Cigarran, M.R. Caballin, H. Braselmann, M. Ribas, J. Egozcue, L. Barrios, Comparison of X-ray dose– response curves obtained by chromosome painting using conventional and PAINT nomenclatures, Int. J. Radiat. Biol. 75 (1999) 1557–1566. [12] F. Darroudi, J. Fomina, M. Meijers, A.T. Natarajan, Kinetics of the formation of chromosome aberrations in X-irradiated human lymphocytes, using PCC and FISH, Mut. Res. 404 (1998) 55–65. [13] P. Virsik, D. Harder, I. Hansmann, The RBE of 30 kV Xrays for the induction of dicentric chromosomes in human lymphocytes, Radiat. Environ. Biophys. 14 (1977) 109– 121. [14] D. Frankenberg, K. Kelnhofer, K. Bär, M. FrankenbergSchwager, Enhanced neoplastic transformation by mammography X-rays relative to 200 kVp X-rays: indication for a
172
H. Schröder, A. Heimers / Mutation Research 517 (2002) 167–172
strong dependence on photon energy of the RBEM for various end points, Radiat. Res. 157 (2002) 99–105. [15] E. Schmid, M. Bauchinger, LET dependence of dicentric yields in human lymphocytes induced by low doses of sparsely ionizing radiations and its implication for risk assessments, Health Phys. 74 (1998) 719–721. [16] D. Harder, D. Regulla, E. Schmid, D. Frankenberg, RBW locker ionisierender Strahlungen: Hat ihre LET-Abhängigkeit Bedeutung für den Strahlenschutz, Strahlenschutzpraxis 2 (2001) 23–31.
[17] T. Straume, High-energy gamma rays in Hiroshima and Nagasaki: implications for risk and w R , Health Phys. 69 (1995) 954–956. [18] J. Weber, W. Scheid, H. Traut, Biological dosimetry after extensive diagnostic X-ray exposure, Health Phys. 68 (1995) 266–269. [19] H.v. Boetticher, Strahlenschutz in der Röntgendiagnostik, Die Röntgenverordnung, Zentralkrankenhaus Links der Weser, Bremen, 1999.
Chromosome aberrations induced in human lymphocytes by in vitro and in vivo X-rays Heike Schröder a,b,∗ , Anna Heimers a,b a
b
Center of Environmental Research and Technology (UFT), University of Bremen, Leobener Strasse, 28359 Bremen, Germany Bremen Institute for Prevention Research, Social Medicine, and Epidemiology (BIPS), Linzer Strasse 8, 28359 Bremen, Germany Received 16 October 2001; received in revised form 14 March 2002; accepted 18 March 2002
Abstract A dose–effect curve is presented obtained by analysis of dicentric chromosomes and centric ring chromosomes in lymphocyte metaphase spreads of three healthy volunteers after in vitro 100 kV X-ray-irradiation of peripheral blood samples. This calibration curve follows a linear quadratic equation, y = c + αD + βD 2 , with the coefficients: y = (0.0005 ± 0.0001) + (0.0355 ± 0.0066)D + (0.0701 ± 0.0072)D 2 . The model is based on 13.231 first-division metaphases analyzed after in vitro exposure to doses ranging from 0.1 to 2.0 Gy at a dose rate of 0.4 Gy min−1 . Significant overdispersion of the observed chromosomal aberrations was evident for dose points 1.0 and 2.0 Gy, respectively. The calibration curve was applied to derive equivalent whole body doses of three subjects after suspected extensive exposure to diagnostic X-rays. © 2002 Elsevier Science B.V. All rights reserved. Keywords: 100 kV X-rays; Dose–effect curve; Human lymphocytes; Chromosome aberrations; Dicentric chromosomes; Biological dosimetry
1. Introduction The analysis of dicentric chromosomes and centric ring chromosomes is considered the most sensitive biological method to quantify exposure to radiation. In the absence of physical measurements biological dosimetry is the method of choice to prove or exclude whether an individual or a group had been affected by ionizing radiation in the past. To accurately quantify the individually absorbed dose it is a prerequisite to establish a representative calibration curve in a standardized in vitro experiment [1]. In medical diagnostics X-rays are applied most frequently with energies of about 100 kV. Therefore, we ∗
Corresponding author. Tel.: +49-421-218-7617; fax: +49-421-218-7616. E-mail address: [email protected] (H. Schröder).
decided to perform this calibration experiment with 100 kV X-rays in such a way that it should enable us to present a dose–effect curve applicable to diagnostic radiography. Chromosome aberration analysis has also been done in three persons who were suspected of an extensive previous exposure to diagnostic X-rays. Based on the calibration curve their equivalent whole body doses were derived from the aberration frequencies observed.
2. Materials and method 2.1. Irradiation conditions Peripheral blood samples were taken by venipuncture from three healthy volunteers in the radiological practice of Dr. Gerhard Schneider and co-workers,
1383-5718/02/$ – see front matter © 2002 Elsevier Science B.V. All rights reserved. PII: S 1 3 8 3 - 5 7 1 8 ( 0 2 ) 0 0 0 6 7 - 0
168
H. Schröder, A. Heimers / Mutation Research 517 (2002) 167–172
Wuppertal, Germany, where the 100 kV X-ray source, RT 100 Müller, (0.4 Gy min−1 , 8 mA, 2.0 mm Al) was located. If not involved in the irradiation procedure, blood samples were kept at 37 ◦ C in a waterbath all the time prior to culture. The irradiation of the samples took place after exact positioning of the probes in an acrylic–glass phantom, which simulated a tissue depth of 4.5 cm. Doses of 0.1, 0.2, 0.5, 1.0 and 2.0 Gy were applied to the whole blood samples. 2.2. Culture conditions Due to pragmatic requirements (irradiation could just be performed after the routine work of the practice had finished and the samples had to be transferred from Wuppertal to Bremen) 18 h after irradiation lymphocyte cultures were set up in the Bremen laboratory according to a standardized cell cycle controlling method described elsewhere [2]. This includes the use of Bromodeoxyuridin which enables the analysis of exclusively first division metaphases. Harvesting of lymphocyte metaphases followed after 48 h culture time. 2.3. Cytogenetic analysis After Fluorescence plus Giemsa staining, chromosome aberration analysis was applied to only first division metaphases. Collection of metaphases was facilitated by an automatic computerized system including a tool of data management (MetaSystems, Altlussheim, Germany). According to our laboratory standards only complete metaphases with 46 centromers were included in the analysis. To analyze exchanges/translocations all chromosomes were assigned to their respective groups according to the International System for Human Cytogenetic Nomenclature [3]. All metaphases were documented and those with suspected chromosomal aberrations were karyotyped routinely for verification. 2.4. Investigated subjects Female, 50 years old, suspected an overdose of X-rays after three diagnostic thorax radiograms within 3 months. After the last one she went under medical care and a radiation dermatitis was supposed. To eval-
uate this diagnosis a blood sample was taken for biological dosimetry 3 months after the last radiogram. Female, 40 years old, also suspected an overdose of X-rays after one diagnostic thorax radiogram because after that she felt ill with various unspecific complaints and a leucopenia had been diagnosed. An independent expert laboratory in the Netherlands calculated a dose of about 600–6000 mGy. 2.5 years later a blood sample has been taken for biological dosimetry to evaluate the radiation dose. Male, 54 years old, had a diagnostic radiation exposure of about 88 mSv, which was documented and had been received during the last 4 years before blood sampling for biological dosimetry. Since this man was in fear of having received an overdose chromosome aberration analysis was performed in order to evaluate the radiation exposure. Dicentric frequencies were compared with the Bremen laboratory control of 0.0005 (95% CI: 0.0003– 0.0008). The equivalent whole body doses were computed based on the in vitro X-ray dose–effect curve. 2.5. Statistics The parameters for the dose–effect curve were obtained by least squares regression, using iteratively reweighted inverse Poisson variances as weights. Goodness of fit was tested with the Pearson χ 2 -value. The confidence intervals were calculated using the method of [4].The variance by mean ratio was used to test for Poisson distribution. Values of (variance/ mean) < 1 indicate underdispersion while values >1 indicate overdispersion. Overdispersion is significant (P < 0.05) when the test quantity u exceeds 1.96 [5,6]. Fisher’s exact test was applied for analyzing differences in dicentric frequencies between investigated subjects and the laboratory control.
3. Results In a total of 13.231 analyzed metaphase spreads 424 dicentric chromosomes, six tricentric chromosomes (tric) and 28 centric ring chromosomes (cR) were found. Counting one tric as two dicentric equivalents, a total of 465 of these unstable chromosome type rearrangements was observed. The percentage
H. Schröder, A. Heimers / Mutation Research 517 (2002) 167–172
169
Table 1 Cytogenetic results of the chromosome aberration analysis obtained after 100 kV X-ray irradiation Dose (Gy)
No. of cells analyzed
No. of dic
No. of cR
Mean dic + cR
95% CI−
95% CI+
No. of excess ace + min
0 0.1 0.2 0.5 1.0 2.0
3013 3014 3020 2063 1333 788
5 22 25 68 110 207
0 0 0 4 7 17
0.0017 0.0073 0.0083 0.0349 0.0878 0.2843
0.00054 0.00457 0.00536 0.02731 0.07259 0.24825
0.00387 0.01105 0.01222 0.04395 0.10519 0.32403
14 49 63 144 227 400
dic: dicentric chromosome; cR: centric ring chromosome; 95% CI−: lower 95% confidence interval (according to the Poisson distribution); 95% CI+: upper confidence interval (according to the Poisson distribution); ace + min: excess acentric fragments and double minutes.
of cR of the total is 6.2%. Table 1 gives the results for the different dose points. Dicentrics and centric rings have been fitted to the linear–quadratic dose effect model: y = c + αD + βD 2 , and it yielded the following coefficients: y = (0.0005 ± 0.0001) + (0.0355 ± 0.0066)D(0.0701 ± 0.0072)D 2 (χ 2 = 6.9; 3 d.f.). Fig. 1 shows the appropriate dose–effect curve with the 95% confidence intervals.The intercellular distribution of the observed dicentric chromosome equivalents and centric ring chromosomes is given in Table 2. The dicentric data show overdispersion after 1.0 and 2.0 Gy. Table 3 presents the results of the chromosome aberration analyses of the three individuals and gives the corresponding doses. Each subject shows a significant increase (P < 0.05) of dicentric chromosomes compared to our laboratory control.
Fig. 1. Dose–effect curve for 100 kV X-rays. The fitted values of the coefficients of the linear quadratic function y = c + αD + βD 2 are: c = (0.05 ± 0.01) × 10−2 ; α = (3.55 ± 0.66) × 10−2 ; β = (7.01 ± 0.72) × 10−2 . The 95% confidence intervals are represented by dashes.
Table 2 Intercellular distribution of dicentric chromosomes and centric ring chromosomes after 100 kV X-ray irradiation Dose (Gy)
Distribution 0
1
0 0.1 0.2 0.5 1.0 2.0
3008 2992 2995 1993 1226 606
5 22 25 68 97 149
∗
2
2 10 25
3
7
4
1
Variance/ mean 0.999 0.994 0.992 1.021 1.084 1.182
u
– – 0.68 2.18∗ 3.61∗
Overdispersion is significant, P ≤ 0.05.
4. Discussion In vitro calibration curves are prerequisites to quantify absorbed doses in exposed individuals or groups. Different dose–effect curves after low LET radiation show that the intercellular distribution of dicentric chromosomes follows a Poisson distribution [7–10]. However, the dose–response curve for 100 keV X-rays presented here shows a significant overdispersion in two of five doses in fact after 1 and 2 Gy. The coefficients agree well with other laboratories [8–11]. There is no full explanation of the observed overdispersion. Unlike other working groups in this experiment overdispersion was observed at the highest doses. Even if tricentric chromosomes and centric ring chromosomes are excluded from the analysis (like other laboratories do) overdispersion still remains statistically significant for both dose points. As the irradiation conditions in our experiment, i.e. dose, dose-rate, radiation quality, and temperature are comparable to others, these factors may not attribute to the observed
170
H. Schröder, A. Heimers / Mutation Research 517 (2002) 167–172
Table 3 Investigated subjects Subject
Age (years)
Time of blood sampling
No. of analyzed cells
No. of dicentrics, mean/cell (95% CI)
Equivalent whole body dose, Gy (CI)
Female
50
3 months after irradiation
1028
3 0.003 (0.0006–0.0085)
0.063 (0.002–0.195)
Female
55
2.5 years after irradiation
1008
5a 0.005 (0.0016–0.0116)
0.105 (0.024–0.250)
Male
54
Immediately after 4 years of diagnostic X-ray radiation exposure
802
4
0.105
0.005 (0.0016–0.0116)
(0.024–0.250)
a
One cell with two dicentrics.
phenomenon. A difference to other published curves is the fact that our cultures had been established around 18 h after the irradiation procedure. But according to Darroudi et al. [12] dicentric chromosomes are formed immediately after irradiation and do not change with time after X-ray doses up to 2.0 Gy. Therefore, the prolonged time between irradiation exposure and culture may be excluded as the cause of the observed deviation from the Poisson distribution. However, this experiment was conducted using the tube voltage of 100 kV instead of the most commonly used 220 kV, and higher effects like increased values of the relative biological effectiveness (RBE) [13,14] were observed after irradiation with soft X-rays. We suppose that the applied 100 kV X-ray irradiation might be a possible reason for the observed overdispersion. By reanalysis the data of the cobalt-60 gamma dose– response curve obtained in our lab with the iteratively reweighted least squares method the linear–quadratic equation has the following coefficients: y = 0.0004 ± 0.0001+0.0064±0.0040D +0.0480±0.0048D 2 . According to Schmid and Bauchinger [15] this method is reliable for evaluating biological effects after low-LET radiation. The ratio of the α-coefficients gives an RBE for low dose X-rays of 5.55±3.62 compared to Co-60 gamma which is significant at the 5% level. Hence the linear term increases with decreasing photon energy. These data show an effect which is discussed at present in terms of radiation protection [16]: the different biological effectiveness of various low-LET radiations. The evaluation of experimental data on the
induction of dicentric chromosomes by photons and electrons yielded a factor of about 6 decrease in effectiveness between Cr X-rays and 3 MeV electrons [15]. According to Straume [17], the use of the single risk and WR or Q-value 1 for all low LET radiations may be inappropriate. In our experiment, acentrics show overdispersion in four of five doses (data not shown). Lloyd et al. [10] also found overdispersion in acentric data after X-ray irradiation. Schmid et al. [8] observed overdispersion for acentrics in five of nine doses in one experiment after X-ray irradiation but Poisson distribution in another experiment. The difference was due to variances in the filter combinations. Schmid et al. [8] pointed at the influence of the X-ray spectrum due to filter conditions applied and concluded that the differences in the dose–effect curves appear to be mainly a consequence of different X-ray bremsstrahlung spectra. This may also be an explanation for the observed overdispersion of dicentric chromosomes after 1 and 2 Gy. Studies on the induction of chromosome aberrations by diagnostic X-rays have been done by Weber et al. [18]. They determined dicentric yields per cell in the range of 0.0037–0.022 (mean: 0.0095) for five patients after extensive exposure. This was an unexpectedly high yield of dicentrics but the authors explained most of the findings with the additional use of iodized radioactive contrast media and the application of computed tomography. All investigated subjects in the present study show a significant elevation of dicentrics compared to the
H. Schröder, A. Heimers / Mutation Research 517 (2002) 167–172
Bremen laboratory control. The dicentric yield of 0.003 and the mean equivalent whole body dose of 63 mGy in subject 1 confirmed the suspicion about a radiation dermatitis. Also the suspected overdose of X-rays allegedly after only one diagnostic thorax radiogram could be verified by biological dosimetry. A dicentric yield of 0.005 in subject 2 leads to a mean equivalent whole body dose of 105 mGy. The finding of one cell with two dicentrics shows the influence of partial body exposure. The dose range of 24–250 mGy has to be regarded as a minimal range because the halflife of lymphocytes and the decline of dicentric chromosomes with time has not been considered. A calculated dose of at least 600 mGy by an independent expert laboratory therefore is not unlikely. Subject 3 is of special interest because his X-ray exposure was well documented over the last 4 years before blood sampling and amounted a value of approximately 88 mGy. The observed dicentric yield of 0.005 gives a mean equivalent whole body dose of 105 mGy which agrees well with the computation based on the X-ray documentation. The dicentric frequencies in the first two cases show that both subjects had overexposures of diagnostic X-rays which were beforehand confirmed by medical reports. These two cases show that minimum requirements in radiation protection had not been fulfilled. An incorrect distance from the X-ray machine to the patient for example can lead to 10 times higher patient doses [19]. Diagnostic radiography requires safe and modern equipment and competent personnel as well as an unequivocal medical indication. Subject 3 has no overexposure from X-rays, the equivalent whole body dose, derived from the dicentric frequency, corresponds with the applied state of the art X-ray doses. In this case the different X-ray exposures given over 4 years could be detected by means of biological dosimetry, indicating that this is still the method of choice also in the low dose range.
Acknowledgements The authors are grateful to Gerhard Schneider and his staff for the opportunity to use their X-ray machine in order to conduct the in vitro irradiation series. They thank Herbert Braselmann and Achim Kranefeld
171
for statistical advice, Ingeborg Grell-Büchtmann, Bettina Dannheim and Boris Oberheitmann for their laboratory work, and Heiner v. Boetticher, Wolfgang Hoffmann and Inge Schmitz-Feuerhake for valuable discussion. References [1] M. Bauchinger, E. Schmid, S. Streng, J. Dresp, Quantitative analysis of the chromosome damage at first division of human lymphocytes after 60 Co ␥-irradiation, Radiat. Environ. Biophys. 22 (1983) 225–229. [2] A. Heimers, H. Schröder, E. Lengfelder, I. Schmitz-Feuerhake, Chromosome aberration analysis in aircrew members, Radiat. Prot. Dosim. 60 (1995) 171–175. [3] ISCN, In: F. Mitelman (Ed.), An International System for Human Cytogenetic Nomenclature, S. Karger, Basel, 1995. [4] C.J. Clopper, E. Pearson, The use of confidence or fiducial limits illustrated in the case of binomial, Biometrica 26 (1934) 404–413. [5] R.C. Radhakrishna, J. Chakravarti, Some small sample tests of significance for a Poisson distribution, Biometrics 12 (1956) 264–282. [6] J.R.K. Savage, Sites of radiation induced chromosome exchanges, Curr. Top. Radiat. Res. 6 (1970) 129–194. [7] A.A. Edwards, D.C. Lloyd, R.J. Purrott, Radiation induced chromosome aberration and the Poisson distribution, Radiat. Environ. Biophys. 16 (1979) 89–100. [8] E. Schmid, M. Bauchinger, S. Streng, U. Nahrstedt, The effect of 220 kVp X-rays with different spectra on the dose response of chromosome aberrations in human lymphocytes, Radiat. Environ. Biophys. 23 (1984) 305–309. [9] L. Fabry, A. Leonard, A. Wanbersie, Induction of chromosome aberrations in G0 human lymphocytes by low doses of ionizing radiations of different quality, Radiat. Res. 103 (1985) 122–134. [10] D.C. Lloyd, A.A. Edwards, J.S. Prosser, Chromosome aberrations induced in human lymphocytes by in vitro acute Xand gamma radiation, Radiat. Prot. Dosim. 15 (1986) 83–88. [11] J.F. Barquinero, S. Cigarran, M.R. Caballin, H. Braselmann, M. Ribas, J. Egozcue, L. Barrios, Comparison of X-ray dose– response curves obtained by chromosome painting using conventional and PAINT nomenclatures, Int. J. Radiat. Biol. 75 (1999) 1557–1566. [12] F. Darroudi, J. Fomina, M. Meijers, A.T. Natarajan, Kinetics of the formation of chromosome aberrations in X-irradiated human lymphocytes, using PCC and FISH, Mut. Res. 404 (1998) 55–65. [13] P. Virsik, D. Harder, I. Hansmann, The RBE of 30 kV Xrays for the induction of dicentric chromosomes in human lymphocytes, Radiat. Environ. Biophys. 14 (1977) 109– 121. [14] D. Frankenberg, K. Kelnhofer, K. Bär, M. FrankenbergSchwager, Enhanced neoplastic transformation by mammography X-rays relative to 200 kVp X-rays: indication for a
172
H. Schröder, A. Heimers / Mutation Research 517 (2002) 167–172
strong dependence on photon energy of the RBEM for various end points, Radiat. Res. 157 (2002) 99–105. [15] E. Schmid, M. Bauchinger, LET dependence of dicentric yields in human lymphocytes induced by low doses of sparsely ionizing radiations and its implication for risk assessments, Health Phys. 74 (1998) 719–721. [16] D. Harder, D. Regulla, E. Schmid, D. Frankenberg, RBW locker ionisierender Strahlungen: Hat ihre LET-Abhängigkeit Bedeutung für den Strahlenschutz, Strahlenschutzpraxis 2 (2001) 23–31.
[17] T. Straume, High-energy gamma rays in Hiroshima and Nagasaki: implications for risk and w R , Health Phys. 69 (1995) 954–956. [18] J. Weber, W. Scheid, H. Traut, Biological dosimetry after extensive diagnostic X-ray exposure, Health Phys. 68 (1995) 266–269. [19] H.v. Boetticher, Strahlenschutz in der Röntgendiagnostik, Die Röntgenverordnung, Zentralkrankenhaus Links der Weser, Bremen, 1999.