NO9800032 StralevernRapport 1998:1

Immunological effects of low dose radiation.

Jon B. Reitan, OveJ. Mellbye, Tone D. Bergan, Per Strand

''''•••^s^^^r^^^i Statens 29-24 stmkvern Norwegian Radiation Protection Authority Postboks 55 • N-1345 0steras • Norway PRINTING ERROR

By a mistake the report subheading has not been printed on the cover. The correct reference and title of the report is:

StralevernRapport 1998:1 Immunological effects of low dose radiation.

Absent or minor effects of Chernobyl fallout in Norway?

Jon B. Reitan, Ove J. Mellbye, Tone D. Bergan, Per Strand Reference: Reitan JB, Mellbye OJ, Bergan TD, Strand P: Immunological effects of low dose radiation. Absent or minor effects of Chernobyl fallout in Norway? StralevernRapport 1998:1. 0steras: Norwegian Radiation Protection Authority, 1998. Language: English.

Key words: Immunology, radiation, radionuclides, Chernobyl, wholebody counting, dose, diet.

Summary: Immunological parameters were compared to annual internal dose from radiocaesium in selected individuals with fairly high levels due to Chernobyl fallout. No or minor derangements were traced, indicating that no serious immunological health effects have to be expected. Referanse: Reitan JB, Mellbye OJ, Bergan TD, Strand P: Immunological effects of low dose radiation. Absent or minor effects of Chernobyl fallout in Norway? StralevernRapport 1998:1. 0steras: Statens stralevern, 1998. Sprak: Engelsk.

Emneord: Immunologi, straling, radionuklider, Tsjernobyl, helkroppstelling, ernaering.

Resyme: Immunologiske parametre ble sammenholdt med interne arsdoser fra radiocesium i utvalgte personer med litt h0ye nivaer etter Tsjernobylnedfallet. Ingen eller bare sma forandringer ble funnet, noe som tyder pa at man ikke kan forvente alvorlige immunologiske helseeffekter. Godkjent/Approved:

Jon B.Reitan, avdelingsoverlege/ Medical Director

Reitan JB, Norwegian Radiation Protection Authority, N-1345 0STERAS Mellbye OJ, Institute of Immunology and Rheumatology, Rikshospitalet, N-0027 OSLO Bergan TD, Norwegian Radiation Protection Authority, N-1345 0STERAS Strand P, Norwegian Radiation Protection Authority, N-1345 0STERAS 20 sider Utgitt 1998-01-15 Opplag 200 Form omslag: Graf, Oslo.Trykk: Jebsen trykk og kopi A/S, 0steras

Bestilles fra: Statens stralevern, Postboks 55, 1345 0STERAS Telefon 67 16 25 00, Telefax 67 14 74 07

ISSN 0804-4910 CONTENT

P Abstract 1 Sammendrag (In Norwegian) 1 Preface 1 Acknowledgments 1 Introduction 2 Material and methods 2 Dosimetry 2 Subpopulations and mitogenic stimulation of mononuclear cells 3 Quantification of immunoglobulin classes and complement factors in serum 3 Antibody testing in serum 4 Controls 4 Correlation studies 4 Results 5 Doses 5 Immunological parameters 5 Correlation studies 8 Discussion 11 Dosimetry 11 Immunological effects 12 Conclusion 14 References 15 ABSTRACT

Immunological parameters were registered in selected Norwegian individuals assumed to be exposed above average to fallout radionuclides from the Chernobyl accident. The doses from internal contamination with 134Cs and l37Cs were assessed by wholebody counting, and ranged from about 0.01 to 10 mSv year"1. The immunological parameters stayed in general within normal limits, but showed nevertheless some correlation to dose. The significance of this finding is discussed, but it is concluded that the Chernobyl accident did not influence the immune system in the population in a serious manner.

SAMMENDRAG

Immunologiske parametre ble registrert hos utvalgte norske personer antatt eksponert mer enn gjennomsnittlig for radioaktivt nedfall fra Tsjernobylulykken. Interndosene fra I34Cs og 13 Cs ble bestemt ved sakalt helkroppstelling av kroppsinnholdet, og varierte mellom omkring 0,01 og 10 mSv ar"1. De immunologiske parametre var stort sett innen normalomradet, men viste allikevel en viss korrelasjon med dosen. Betydningen av dette funnet diskuteres, men det konkluderes at Tsjernobylulykken ikke influerte befolkningens immunsystem pa noen alvorlig mate.

PREFACE

This study was made some years ago. Interpretations and discussions about the findings (or lack of findings) have later been numerous, and journal referee opinions have diverged. It is a study based on possibilities and available material during follow-up of the Chernobyl accident in Norway. It is no extensive and thoroughly planned investigation, as resources for such health studies have grossly been lacking. Therefore the design, sampling and measurement strategy may be criticized, and critics may even be valid.

Nevertheless, it is the opinion of the authors that the data should be published, not least because the measured individuals should have a right to know what was really found. We have therefore decided to publish the data as a report, instead of a paper in a refereed journal, in order not to delay its publication further. Individual data have been communicated to the participants separately, and we regret the delay.

ACKNOWLEDGMENTS

This project was supported by the Norwegian Food Control Authority. We are highly indebted to the technical staff at the institutes. Ingar Amundsen MSc processed the data, and Tor Haldorsen MSc helped with the statistics. INTRODUCTION

After the Chernobyl accident in 1986, radioactive deposits were traced in most European countries. Even in Norway, fairly high levels of contamination occurred in parts of the country (Backe et al 1987), in spots up to 600 kBq/m2 of radiocaesium. The doses to the Norwegian population have been minimal, except for some rural populations living of reindeer breeding and mountain sheep production (Strand et al 1989, Strand et al 1992a). Shortly after the accident, widespread public anxiety developed (Weisaeth 1990), probably most serious in the regions with highest pollution (Reitan 1990).

Possible immunological effects represented one aspect of the public concern, because it is well known that ionizing radiation in high or medium doses may derange the immune system (Anderson & Warner 1976). Immunological effects have also been registered in persons acutely exposed in Chernobyl during the accident itself (Kalinin et al 1991, Kozyreva et al 1990). Effects of low doses are more uncertain. In atomic bomb survivors, age- and dose-dependant effects have been shown for fairly low doses (Akiyama et al 1989, Kusunoki et al 1988). Effects have also been shown in the Chernobyl fallout areas in the former Soviet Union (Galitskaya et al 1990, Hellebostad et al 1990, Orodovskaya et al 1995, Titov et al 1995, Yarilin et al 1995). However, there is a lack of observations on populations continously exposed for years to artificial nuclides in the mSv range appropriate for risk evaluation of the Chernobyl fallout.

The Norwegian Radiation Protection Authority has carried out whole-body measurements of radioactive caesium in selected populations since the bomb tests in the 60ies (Strand et al 1989, Strand et al 1992a, Westerlund et al 1987), and it was regarded feasible to correlate immunological parameters with individual internal dose determinations. After contact with organizations of the affected populations, this pilot study was therefore started. Some of the individuals were also studied with cytogenetics, and the results have been reported previously (Br0gger et al 1996).

MATERIAL AND METHODS

Wholebody caesium measurements of individuals were carried out according to previous programs with wholebody counting in three different municipalities with fairly high levels of contamination. Most of the persons measured in 1991 also consented in blood sampling, and the number of persons in this material therefore comprise 49 from Snasa, 9 from R0ros and 18 from 0ystre Slidre, in total 76 individuals. Blood samples were transported as fast as possible to the immunological institute. The transport time from Snasa was about 5 hours. Analyses of cellular immunology were therefore not done for these individuals. The transport time from R0ros by plane and car was 2 hours, and from 0ystre Slidre by car 3,5 hours.

Dosimetry

The individuals were measured using an 3" Nal detector connected to a Canberra Series S30 multichannel analyzer (Canberra, Connecticut, USA) as previously described 2 (Strand et al 1990). The doses per year were calculated based on body content of 134Cs and 137Cs, using standard dose conversion factors recommended by the International Commission on Radiological Protection (ICRP 1979). Most persons measured in 1991 had also been measured previous years since 1987, making it possible both to analyse effects of cumulated doses and doses last year. Some missing values for persons not measured annually since 1987 had to be calculated by inter- and extrapolation. No correction for small residual amounts of l37Cs originating from the athmospheric atomic bomb tests were attempted.

Subpopulations and mitogenic stimulation of mononuclear cells

Blood was drawn with heparin. Immediately after arrival at the immunological laboratory, mononuclear cells (MNC) were separated on Isopaque-Ficoll. MNC from healthy laboratory personnel were always included in the same set-up.

The subpopulations were measured by flow cytometry, using a FacScan analyzer and mouse monoclonal antibodies (Table I) from Becton Dickinson (Mountain View, California). The relative number of monocytes, total T-cells, total B-cells, helper T- cells, and cytotoxic/«suppressor» T-cells was measured. The sum of the results for the first three populations were considered as the total amount of classifiable cells, and the results for all the subpopulations were expressed as a percentage of that sum.

TABLE I: Monoclonal antibodies used for identification of subpopulations of mononuclear cells

Antibody Subpopulation CD14(Anti-Leu-M3)FITC' Monocytes CD3 (Anti-Leu-4) FITC Total T-cells CD8 (Anti-Leu-2a) FITC Cytotoxic/"suppressor" T-cells CD4 (Anti-Leu-3a) FITC Helper/inducer T-cells CD 19 (Anti-Leu-12) FITC Total B-cells

1: FITC = Fluorescein isothiocyanate

Mitogenic stimulation of MNC was performed for 72 hours with phytohemagglutinin (PHA), concanavalin A (Con A), and pokeweed mitogen (PWM). During the last 18 hours the cultures were pulsed with 3H-thymidine, and the amount of incorporated thymidine was measured in a [3-scintillator. The results were expressed as the stimulatory index (SI) - i.e. cpm in culture with mitogen / cpm in culture without mitogen.

Quantification of immunoglobulin classes and complement factors in serum

IgG, IgA, IgM, and the complement factors C3 and C4 were quantified with use of a 3 nephelometer and antisera from Behringwerke AG (Marburg, Germany).

Antiantibodv testing in serum

For screening for antinuclear antibodies (ANA), sera were tested by indirect immunofluorescence technique, using the human HEp-2 cell line as substrate and a FITC-conjugated rabbit anti-IgG antiserum from DAKO A/S (Glostrup, Denmark). Positive sera were further tested for anti-ds (double stranded) DNA antibodies by indirect immunofluorescence technique with Crithidia luciliae as substrate, and for antibodies to extractable nuclear antigens (ENA) with double diffusion technique.

Waaler-Rose test for rheumatoid factor (RF) was performed by passive hemagglutination technique, using human erythrocytes coated with rabbit IgG antibodies. Latex RF test was done with a commercial kit (NA Latex, Behringwerke AG, Marburg, Germany). This assay was based on nephelometry, using polystyrene particles coated with human IgG.

Antibodies to gastric parietal cells, smooth muscle, mitochondriae, glomerulus basal membrane, adrenal cortex cells and neutrophil cytoplasmic antigen (ANCA) were tested by routine indirect immunofluorescence tests with FITC-labelled anti-IgG or anti- F(ab')2 antisera from DAKO A/S (Glostrup, Denmark). Antibodies to thyroglobulin and microsomal thyroid antigen (thyroid peroxidase) were tested by passive hemagglutination technique, using kits from Murex (Dartford, England), whereas anti-ss (single stranded) DNA was tested by ELISA technique.

Controls

Testing of 80 healthy blood donors was performed to establish reference values (2.5 and 97.5 percentiles) for the immunoglobulin and complement quantifications and cut-offs (5%) for the autoantibody testings.

For the cellular immunology studies one blood sample from healthy laboratory personnel was included in each set-up as a control that the systems worked. Reference ranges were obtained by testing in the laboratory of a group of 18 healthy hospital employees. Since the values obtained were clearly assymetrically distributed, log- normal transformation of the data was performed. Geometric mean ± 2 SD logarithms were retransformed to reference limits by their antilogs. For the mitogen stimulations, the stimulatory index (SI) for the controls should be at least 25 for PHA and 5 for ConA and PWM.

Correlation studies

The immunological data were correlated to the doses, in some instances also to the age of the individuals, using Statgraphics 5.0 software (STSC, Rockville, Maryland). Linear regressions and Spearman's rank correlations were mostly applied, deviations from this are mentioned in the text and tables. RESULTS

Doses

76 individuals were measured with wholebody counting in 1991. The number of previous measurements since 1987 should therefore ideally be 304, however, only 166 measurements had been done. The missing 138 values were calculated by inter- and extrapolation. The doses based on body content of 134Cs and 137Cs is shown in Table II together with the estimated mean doses to inhabitants of the municipalities in question (Strand et al 1992a).

Immunological parameters

For the cellular immunological parameters, most of the values observed for the various lymphocyte subpopulations were within the ranges established in our laboratory (Table III). There was no increased frequency of individuals with increased relative concentration of total and helper T-cells, nor with decreased concentration of B-cells (/ test). For the mitogen stimulations, all the individuals in the study group gave a significant normal response to PHA and ConA. With PWM very low stimulation was observed both for the individuals studied and the normal controls in the various set-ups. These results were therefore considered to be technical failures.

TABLE II: Internal doses in mSv pr. year in 1991 and for the years 1986-1991 in selected groups of Norwegians. Values for the study group based on individual measured body content of 134Cs and l37Cs, and for the municipalities calculated total mean based on food contamination in the same municipality.

Study group Municipality Study group mean Study group range mean max. min. 1991 Snasa + R0ros O.O93 0.73 2.17 0.11 0ystre Slidre O.133 0.54 2.04 0.01 1986-1991 Snasa + R0ros 0.104 1.08 (1.15) 12.985 0.13 0ystre Slidre6 0.164 0.61 (0.64) 2.29 0.10

1: Calculated weighted mean. Ref.: Strand et al.: Health Phys 62:512-518,1992 2: Mean of all measurements. In brackets: Mean of means for the individual years 3: Data for 1989 4: Data for 1986-1989 5: Highest recorded individual value in Snasa in 1988 6: No wholebody measurements performed in 1990 TABLE III: Frequency of individuals with relative number of lymphocyte subpopulations (percentage of sum of macrophages, B-cells and T-cells) outside range in reference group (Individual increased/ decreased values in footnotes).

Parameter tested Reference range Frequency of Frequency of (per cent) decreased values increased values B-cells 2-11 1/201 0/20 Total T-cells 62- 100 0/20 0/20 Helper T-cells 31 -72 1/2O2 0/20 Cytotoxic/" suppressor" 18-66 2/2O3 1/204 T-cells CD4+/CD8+ cell ratio 0,6 - 3,4 1/205 2/206

1: 0 %9 2: 26 % 3: 13, 16 % 4: 66 % 5: 0.5 % 6: 3,8, 4,5 %

TABLE IV: Frequency of individuals with immunoglobulin or complement concentrations outside reference limits (Individual increased/ decreased values in footnotes).

Parameter tested Reference values Frequency og Frequency of (g/1) decreased values increased values IgG 8.0- 17.0 0/76 4/76' IgA 1.0-5.5 1/762 1/763 IgM 0.7-3.8 2/764 1/765 C3 0.5-1.3 0/76 0/76 C4 0.1 -0.5 0/76 0/76

1: 18.3, 18.1, 19.4, 19.7 g/1 2: 0.8 g/1 3: 5.7 g/1 4: 0.4, 0.6 g/1 5: 4.4 g/1

For the quantifications of immunoglobulin classes and complement factors, and for the autoantibody testings, almost all the individual results were within the reference limits established in the laboratory performing the analyses. The frequency of values outside the reference limits are shown in tables IV and V, and for none of the parameters listed was there a statistically significant difference between the study group and the controls 6 (X test). The individual results for some of the parameters are included in the figures.

TABLE V: Frequency of individuals with autoantibody values above upper reference limit (Values in titers if not otherwise stated. Individual increased/decreased values in footnotes).

Autoantibody tested Upper reference limit Frequency of increased values Controls Study Group Rheumatoid factor Waaler's test <32 4/80 2/76' Latex test <25 U/ml 1/80 3/762 Antinuclear factors ANA-test <32 2/80 9/763 Anti-ssDNA < 220% 2/80 3/764 Anti-gastric parietal cells < 16 1/80 3/765 Anti-smooth muscle < 16 3/80 1/766 Anti-mitochondriae < 16 1/80 0/76 Anti-thyroglobulin <20 4/80 1/767 Anti-thyroid perox. (TPO) <100 7/80 1/768 ANCA < 16 0/80 0/76 Anti-glom. basal membrane <2 0/80 0/76 Anti-adrenal cortex cells <2 0/80 0/76

1: 128, 16400 2: 69, 84, 2230 U/ml 3: 32, 32, 32, 64, 64, 64, 256, 1024, >1024. All these sera were negative for anti-ds (double stranded) DNA and anti-EENA (extractable nuclear antigens). 4: 229, 314, 332% 5: 256, 1024, 1024 6: 16 7: 80 8: 400 9: 64 (perinuclear)

As shown in table V, some of the individuals in the study group had high titers of autoantibodies usually associated with connective tissue disorders. These individuals were mainly from Snasa and of Sami ethnic origin. Their local physician was contacted, but in none of these cases were symptoms or signs of connective tissue disorders revealed. There seemed to be no association between these increased values and a high radiation dose (data not shown).

Correlation studies Correlation studies

The correlation studies in general showed that the same correlations or lack of correlations were found whether using the 1991 dose alone or the cumulated dose. Because 1991 doses probably are most relevant to correlate to 1991 immunological data, we report only these data here.

The cellular parameters showed a negative correlation between CD4+-cells (helper-T) and dose with regression coefficient of-0.37 but with p=0.11 (Data not shown). CD8+-cells (cytotoxic/"suppressor"-cells) correlated positive with linear correlation coefficient 0.40 and p=0.083 (Data not shown). CD4+/CD8+ ratio (helper/suppressor-ratio) showed by reciprocal regression (l/Y=a+bx) some correlation to dose with correlation coefficient 0.48 and p=O.O33, but in a linear model the correlation coefficient was -0.32 and p=0.17 (Data not shown). CD4+, CD8+ and their ratios were also compared to age, but showed no linear correlation (Data not shown). Total B-cells did not correlate with dose, but an impression of a bellshaped function may arise (Fig 1). 10 F o :

*" •*• o 8 A" - - 1 \ 1 V ° 1 \ o , \ o t * 1 ° • 4 • o o -- ^_ o - •i • coo o o -I 2 - o o o •

-o 0 1 1 1 1 0 0,4 0,8 1,2 1,6 2 2,4 Dose (mSv year-1) Fig. 1 Per cent B-lymphocytes (CD 19+ cells) of total number classifiable mononuclear cells as a function of 1991 year dose in mSv based on wholebody measurements. Individual values and linear regression (approximative log-normal distribution drawn by hand).

The mitogen stimulation indices showed some dependence on dose, with the lower values associated with the higher doses. When analysed in a multiplicative model the correlation coefficient was -0,46 and p= 0.039 for PHA, -0.50 and p=O.O25 for ConA. (Even the low technically unsuccessful PWM results correlated in the same manner). Also by dividing the material by dose below or above 0,5 mSv, a significant difference was found (Table VI). 8 TABLE VI: Lymphocyte mitogene stimulation index ± SE i relation to 1991 dose.

Parameter tested Dose < 0,5 mSv1 Dose > 0,5 mSv2 SI- ConA3 77,3 ± 20,0 23,7 ± 6,9 SI-PHA4 187,5 ± 40,6 74,6 ± 15,8

1: N=ll 2: N = 9 3: p < 0.04 (t-test) 4: p < 0.03 (t-test)

IgM and IgA showed no significant correlation to 1991 dose (Data not shown). IgG also showed scattered values, but correlated negatively (correlation coefficient -0.26, p=0.026) with dose (Fig. 2). The complement factors C3C and C4 (Figs. 3 and 4) were however positively correlated with dose with correlation coefficients 0.21 (p=0.065) and 0.23 (p=0.044). By use of Pearson's product-moment correlation, C4 had a correlation coefficient of 0.30 with p=0.0074.

1 1 1 1 1— 20 -' 0 o 18 - O ° - 0 o 16 o o° oo o o 0 ° 14 - o° ^ o -u 0^ " 6° -:i > _ o° o 12 o o, o , - ^"' »*- 2 ° ° . L c °o,o ° °o° —i 10 o o i U 8 o i

4 - 2 - - - 0 1 1 1 1 1 0 0,4 0,8 1,2 1,6 2,4 Dose (mSv year"-1 )

Fig. 2 Concentration of immunoglobulin IgG (nephelometry units) as a function of 1991 year dose in mSv based on wholebody measurements. Individual values and linear regression. 0,4 0,8 1,2 1,6 2 2,4 Dose (mSv year-1) Fig. 3 Concentration of complement factor C3C (nephelometry units) as a function of 1991 year dose in mSv based on wholebody measurements. Individual values and linear regression. 0,6 —i 1 1 1 1 1 i

o 0,5 - o

00 0 o 0 0 0,4 o o • 0 o o o o o „ o 0,3 0 O jl—• •—' oo o

o o O 0 o o o 0,2 "~ o o 0 0

0,1 -

0 Hi i i 1 1 1 1~ 0 0,4 0,8 1,2 1,6 2 2,4 Dose (mSv year1) Fig. 4 Concentration of complement factor C4 (nephelometry units) as a function of 1991 year dose in mSv based on wholebody measurements. Individual values and linear regression.

10 The ANA, Waaler and latex values did not show any correlation to dose at all, nor did they show any correlation to age.

DISCUSSION

Dosimetry

Wholebody counting gives a point estimate only of the body radiocaesium content during the year. Previous studies on variation of mean doses in community populations from month to month indicate that this variation generally is low (Strand et al 1989). However, the differences during the year in the populations studied here may be greater, because they represent reindeer breeders, mainly of sami origin, (Snasa and R0ros) and mountain farmers (0ystre Slidre) with not completely ordinary diet habits (Strand et al 1992b).

Nor is the dose reported here representing the total dose from radionuclides to the individuals. External radiation from ground deposition is not measured, but is assumed to be more related to the common living conditions of the groups, and not so much to the individuals. Natural internal nuclides as 40K and 2l0Po are excluded, and dose contributions from 90Sr have not been estimated due to very low proportion in the Chernobyl fallout. However, we assume some correlation between the intake of Cs, Sr and other artificial radionuclides, although the correlation may be weak (Evensen et al 1990, Strand et al 1992a).

137Cs originating from the bomb test fallout could be determined based on the known initial ratio of 134Cs to 137Cs. It was, however, not subtracted from the dose estimates, because possible effects were thought to be a result of both cesium origins. But this also implies that the total cumulative dose before the Chernobyl accident is not taken into consideration. After the Chernobyl accident, dietary advice has been given by the health authorities, especially to reindeer breeders and other groups with an expected high intake of radiocaesium. Some persons may have had a comparable high intake before the accident and have reduced their intake. Some people may also have altered their intake at some later point. Despite this possible bias there was found a close correlation between 1991 dose and cumulated 1987-1991 dose in the persons actually measured each year. Persons with high levelse of radioceasium in 1991 had high values earlier, and persons with low values had low values earlier. The method of inter- and extrapolation of missing doses may further imply a close proportionality of the estimated 1991 dose and cumulated dose 1987-1991 for the group in total.

A serious problem relates to the intake of 2l0Po. Through the lichen-reindeer pathway 2l0Po is concentrated in reindeer meat, and contributes significantly to the total dose of reindeer breeders. Problems with this nuclide have been discussed in an earlier report (Evensen et al 1990). 210Po would certainly increase the dose estimates in the counties of Snasa and R0ros, probably not to the same extent in 0ystre Slidre, but has not been taken into account in the present study.

11 We have therefore accepted the Cs-based 1991 doses as a relative dose appropriate for defining a dose-response regression. The absolute doses may have been somewhat higher.

Immunological effects

In general, the immunological results look fairly normal. An obvious problem is that the immunological status is assessed 5 years after the accident. The normal results found in this study does not exclude the possibility that significant effects could have subsided, or that such effects will evolve later. In Byelorussia immunological effects in children shortly after the accident showed later partial normalization (Hellebostad et al 1990, Titov 1994, Titov et al 1995). However, time sequences of immunological effects of continous low-dose irradiation from e.g. fallout have not been detailed reported to our knowledge, but in reindeer breeders in the former Soviet union, persistant depression of T-lymphocytes has been observed (Baarli 1992). Combined effects of irradiation and aging have been observed in atomic bomb survivors (Kusunoki et al 1988). But with high consumption of reindeer meat, age also correlates to radiation dose, both from manmade fallout and from natural sources as e.g. 2l0Po. However, when analyzing immunological effects versus age in the present material, no significant correlations were found.

In vitro medium dose studies of different subtypes of unstimulated human lymphocytes has indicated that CD8+ cells are twice as radiosensitive than CD4+ cells (Stewart et al 1988). However, the lowest dose used by Stewart et al was 250 mGy, and linearity cannot be unequivocally assumed. The relative proportion of the CD8+ and CD4+ cells found in this study is not consistant with such relative radiosensitivity. Neither CD4+ nor CD8+ did correlate to age. Oradovskya et al (1995) found an increase of CD4+ cells and no or only slight effects on CD8+ cells in most of their Russian populations. In vitro radiosensitivity of human B-cells for medium radiation doses seems to be in the same range as the T-cells (Stewart et al 1988). Studies of mouse B-cells has shown them to be much more radiosensitive than T-cells (Anderson et Warner 1975), or quite opposite (Takeuchi et al 1992). In lethal irradiation of man, plasma cells have been observed to be the main residual bone marrow cell type (Reitan et al 1990). Intermitotic death of thymocytes has been shown to be due to apoptosis (Yamada et Ohyama 1988) and it may be that a constant low-dose irradiation from e.g. 90Sr may have induced changes also in the bone marrow function (Stokke et al 1968). The complex interactions of the immune system during a response may imply that a specific cell subpopulation (or a factor produced by one) is limiting at one time or dose level and unimportant at another time or at another dose level. Moreover, the clinical significance of immunosuppression in relation to e.g. carcinogenicity is far from clear (Prehn 1994).

An interesting result in this study is the relation between dose and mitogen stimulation indices, indicating a low dose radiation induced depression. Decreased stimulation by PHA, more dubious by ConA, has been observed in dogs after inhalation of 239PuC>2 (Davila et al 1992). The bone marrow dose from such inhalation must be minimal, and the effects may be of toxic origin. Severe depression of mitogene stimulation by ConA has been found in mice, but after substantial higher doses (Takeuchi et al 1992). Japanese atomic bomb survivors in general showed a moderately depressed PHA 12 response (Akiyama et al 1983). The exposed group (<500 mSv) of the survivors residing in the US have shown increased mitogene response, interferon production and natural cell mediated cytotoxicity (Bloom et al 1987). Studies on survivors still residing in Hiroshima did not confirm this (Kato et al 1987), except for some effect of age on natural killer cell activity (Bloom et al 1988). Yarilin et al (1995) observed a reduced PHA stimulation in Chernobyl cleanup personell, but the doses must have been fairly high. Dose-effect relationships for stimulation indexes are apparently not known, but can obviously be of some clinical significance. We have, however, no indication that this has been the case in our material.

The negative correlation to dose of the IgG in the present material might be explained by a radiation induced reduced number of B (CD 19+) lymphocytes, but by linear regression, no correlation was found for the B-cells. No consistent changes were observed for other globulin classes, and a wide scatter of the IgG values also points against any biological explanation for a dose-effect relationship. Nevertheless, Oradovskaya et al (1995) have reported reduced IgG values in their moderately irradiated cases, whereas Titov et al (1995) report an initial decrease (40-50 days after the accident) with later increase. In contaminated areas in Byelorussia, immmune globulins and total B-cells were increased in children a few years after the Chernobyl accident (Galitskaya et al 1990). In atomic bomb survivors more than 40 years after the bombing a dose dependent increase in IgM and IgA was found, whereas IgG and IgE were unaffected (Fujiwara et al 1994). The latter authors also stressed that radiation dose only could explain about 10% of the variance in the material. In general, therefore , it seems that any consistent reaction pattern have not so long been revealed.

By log-normal curve fitting to our data some consistency of the data on the B cells may appear. Biological models consistent with such data have been forwarded (Oftedal 1991, Burlakova 1995)), but because all values fit nice within the normal range, explanations for such a curve are at best speculative. We therefore consider the data in the present material, both regarding B-cells and IgG, consistent with a no effect result.

A positive correlation was found for complement factors C3C and in particular C4 equally in sami and Caucasians. It should otherwise be mentioned that the frequency of the C3 1 (C3 F) complement gene normally is low in samis, which also lack rare C3 variants, whereas the frequency of the C6 B gene is higher in Kautokeino and Karasjok samis than in other Norwegians (Eriksson 1988). Increases in complement factors C3 and C4 in humans have also been reported from radioactivity contaminated areas in Russia (Vasilev et al 1994). Today, it is unknown whether there is any biological significance of variations within normal reference limits of complement or immunoglobulins. For other serum components, however, it is clear that variations within the normal ranges are of significance.

The frequency of increased titers in autoimmune reactions as ANA and RA may not be significantly different from the control material of the laboratory (Table IV). Suspicious however, especially in the group from Snasa, is that the persons with results above normal show fairly high titers. Autoimmunity has been claimed to have some relation to deficiency of cytotoxic/«suppressor» T-cells. However, CD8+ cells correlated positively with dose in this study. Moreover, the relation between CD8+ cells and suppressor 13 activity is today far from clear (Nossal 1993). High dose irradiation of mice can induce autoimmune disorders (Sakaguchi et al 1994). These seem induced by alterations in the immune system and not by factors in the organs in question, and injection of normal T cells indicated some autoimmune prevention by CD4+ cells but not by CD8+ cells. Significant dose dependant increase in RA has been found in atomic bomb survivors, but not for ANA, antithyroglobulin or anti-thyroid-microsomal autoantibodies (Fujiwara et al 1994). The autoimmune markers in general in our material show no correlation to 1991 dose, so we consider the increased titers in some individuals probably not related to nuclear pollution.

CONCLUSION

In this pilot study of those Norwegian individuals most heavily exposed to the Chernobyl fallout, immunological parameters generally stayed within normal limits. However, some parameters apparently within the assumed normal range did in fact correlate to the estimated individual dose as assessed by wholebody counting of radiocaesium content.

The small possible effects revealed in this study may represent real biological effects, but do not necessarily represent a health detriment. The significance of variations of immunological parameters within the normal range is speculative only at the moment. However, the definition of normal values in medicine is generally meening a statistical group mean with some deviation, not a normative or "ideal" value. In other words: It is not normal to be completely healthy. Biological significance of variations in e.g. erythrocyte sedimentation rate as a prognostic parameter of cancer is highly increased if individual reference values are used instead of the normal range (Hannisdal et al 1993). Sensitivity of scoring of radiation bone marrow effects improved if mice from the same litter were used as controls (Stokke et al 1968). Thus, choice of appropriate control groups may heavily affect the possibility to trace immunological low dose radiation effects in man. Moreover, we can not exclude the possibility that temporary more obvious effects may have existed earlier or may evolve later, and only adult persons have been studied. The results give a motivation for extended and continued studies of possible low dose irradiation effects. However, because the persons studied represent the maximally exposed groups in Norway, it seems reasonable to conclude that the Chernobyl accident did not influence the immune system in the population in a serious manner.

14 REFERENCES

Anderson RE, Warner NL 1975: Radiosensitivity of T and B lymphocytes. Ill: Effect of radiation on immunoglobulin production by B cells. J Immunol 115:161-169

Anderson RE, Warner NL 1976: Ionizing radiation and the immune response. Adv Immunol 24:215-335

Akiyama M, Yamakido M, Kobuke K, Donald SD, Hamilton HB, Awa AA, Kato H 1983: Peripheral lymphocyte response to PHA and T cell population among atomic bomb survivors. Radiat Res 93:573-580

Akiyama M, Zhou OL, Kusunoki Y, Kyoizumi S et al. 1989: Age and dose related alteration of in vitro mixed lymphocyte culture response of blood lymphocytes from A-bomb survivors. Radiat Res 117:26-34

Baarli J (Ed) 1992: Novaya Zemlya. Ecological security of underground nuclear tests. Report 92-03 (Oslo: Department of Physics, University of Oslo)

Backe S, Bjerke H, Rudjord AL, Ugletveit F 1987: The fallout of caesium in Norway after the Chernobyl accident. Radiat Prot Dosim 18:105-107

Bloom ET, Akiyama M, Kusunoki Y & Makinodan T 1987: Delayed effects of low- dose radiation on cellular immunity in atomic bomb survivors residing in the United States. Health Phys 52:585-591

Bloom ET, Akiyama M, Korn EL, Kusunoki Y, Makinodan T 1988: Immunological responses of aging Japanese A-bomb survivors. Radiat Res 116:343-355

Br0gger A, Reitan JB, Strand P, Amundsen I: Chromosome analysis of peripheral lymphocytes from persons exposed to radioactive fallout in Norway from the Chernobyl accident. Mutat Res 361:73-79,1996

Burlakova EB: Low intensity radiation: Radiobiological aspects. Radiat Prot Dosim 62:13-18,1995

Davila DR, Guilmette RA, Bice DE, Muggenburg BA, Swafford DS, 239 Haley PJ 1992: Long-term consequences of PuO2 exposure in dogs: persistent T lymphocyte dysfunction. Int J Radiat Biol 61:123-133

15 Evensen JF, Reitan JB, Westerlund EA, Br0gger A 1990: Cesium-137 body burden and chromosome aberrations in Norwegian lapps 1965-1978. Radiation and cancer risk ed Brustad T, Langmark F, Reitan JB (New York: Hemisphere) pp 21-30.

Eriksson AW 1988: Anthropology and health in lapps. Coll Anthropol (Zagreb) 12:197-235

Fujiwara S, Carter RL, Akiyama M, Akahoshi M, Kodama K, Shimaoka K, Yamakido M 1994: Autoantibodies and immunoglobulins among atomic bomb survivors. Rad Res 137:89-95

Galitskaya NN, Khmelevskaya LA, Zhuk GM, Vorontsova TV 1990: Estimation of the immunological status of the children from the areas with abnormal radiation (In Russian). Zdravoohranenie Belorussii (Minsk) 1990(6):33-35

Hellebostad M, Reitan JB, Salbu B, Vollset SE 1990: Radioactive fallout and its effects in Byelorussia. NLVF-Report 151 (Oslo: NLVF)

Hannisdal E, Gundersen S, Kval0y S, Lindegaard MW, Aas M, Finnanger AM, J0rgensen OG, Theodorsen L, H0st H 1993: Follow-up of breast cancer patients stage I-II: A baseline strategy. Eur J Cancer 29A.992-997

ICRP 1979: Limits for intakes of radionuclides by workers. International commission on Radiological Protection. Publication 30. (Oxford: Pergamon Press)

Kalinin NL, Kukharkova TV, Karpun II, Baranov AE 1991: Dynamics of the immunohematologic indices in patients with acute III-IV degree radiation sickness in the post-pancytopenia period (In Russian). Med Radiol (Moscow) 36:32-36

Kato H, Schull WJ, Awa A, Akiyama M & Otake M 1987: Dose-response analysis among atomic bomb survivors exposed to low-level radiation. Health Phys 52:645-652

Kozyreva TV, Nikiforova NA, Kalmykova II, Skobeltsyna ES, Sorochan PP, Starodubtseva AN 1990: Immunity status in persons participating in liquidation of the effects of the accident at the Chernobyl nuclear power plant (In Russian). Gematol Transfuziol 35:17-19

Kusunoki Y, Akiyama M, Kyoizumi S, Bloom ET, Makinodan T 1988: Age-related alteration in the composition of immunocompetent blood cells in atomic bomb survivors. Int J Radiat Biol 53:189-198

16 Nossal GJV 1993: Tolerance and suppression. Clinical aspects of immunology. 5th edn. ed Lachmann PJ, Peters K, Rosen FS, Walport MJ. (Oxford: Blackwell Scientific Publ) pp 691-708

Oftedal P 1991: Biological low-dose radiation effects. Mut Res 258:191-205

Orodovskaya IV, Fadeeva ID, Ulyanova NV, Chernetsova LF, Nikonova MF, Litvina MM: Six-year observation of immune state of persons affected by the Chernobyl accident. Radiât Prot Dosim 62:63-67,1995

Prehn RT 1994: Stimulatory effects of immune reactions upon the growths of untransplanted tumors. Cancer Res 54:908-914

Reitan JB 1990: Radioactivity and health effects (In Norwegian). NLVF's forskningsprogram om radioaktivt nedfall, ed Gunner0d TB, Garmo TH (Oslo: Informasjon fra Statens fagtjeneste for landbruket) pp 11-20

Reitan JB, Stavem P, Kett K, Hoel PS 1990: The 60Co accident in Norway 1982: A clinical reappraisal. The medical basis for radiation accident preparedness II. ed Ricks RC, Fry SA (New York: Elsevier) pp 3-11.

Sakaguchi N, Miyai K, Sakaguchi S 1994: Ionizing radiation and autoimmunity. Induction of autoimmune disease in mice by high dose fractionated total lymphoid irradiation and uts prevention by inoculating normal T cells. J Immunol 152:2586-2595

Stewart CC, Stevenson AP, Habbersett RC 1988: The effect of low-dose irradiation on unstimulated and PHA-stimulated human lymphocyte subsets. Int J Radiât Biol 53:77-87

Stokke T, Oftedal P, Pappas A 1968: Effects of small doses of radioactive strontium on the bone marrow of the rat. Acta Radiol 7:321-329

Strand P, B0e E, Berteig L, Berthelsen T, Strand T, Trygg K, Harbitz O 1989: Whole-body counting and dietary surveys in Norway during the first year after the Chernobyl accident. Radiât Prot Dosim 27:163-171

Strand P, Selnaes T, Andersson A 1990: Wholebody measurements in South Norway after the Chernobyl accident 1987-1989 (In Norwegian with English summary). SIS-report 1990:4 (0steras: SIS)

Strand P, Selnaes TD, Reitan JB 1992a: Area and time distribution of external and internal doses from the Chernobyl fallout: The lack of correlation in Norway. Health Phys 62:512-518 17 Strand P, Selnaes TD, B0e E, Anderson-S0rlie A 1992b: Internal doses to different parts of the Norwegian population after the Chernobyl accident and the effect of dietary advice. Health Phys 64:385-392

Takeuchi M, Shibata H, Nasu T 1992: Effect of soft X-ray irradiation on immunological functions in mice. J Vet Med Sci 54:653-658

Titov LP, Kharitonic GD, Gourmanchuk IE, Ignatenko SI: Effects of radiation on the production of immunoglobulins in children subsequent to the Chernobyl disaster. Allergy Proc 16:185-93,1995

Titov LP 1994: Personal communication

Vasilev AV, Samsonov MA, Pokrovskii VB, Pokrovskaia GR, Levachev MM 1994: Characteristics of the humoral immune system and lipid peroxidation processes in people living in territories contaminated with radionuclides (In Russian). VoprMedKhim 40:21-24

Weisaeth L 1990: Reactions in Norway to fallout from the Chernobyl disaster. Radiation and cancer risk ed Brustad T, Langmark F, Reitan JB (New York: Hemisphere) pp 149-155

Westerlund EA, Berthelsen T, Berteig L 1987: Cesium-137 body burdens in Norwegian Lapps. Health Phys 52:171-177

Yamada T, Ohyama H 1988: Radiation-induced interphase death of rat thymocytes is internally programmed (apoptosis). Int JRadiat Biol 53:65-75

Yarilin AA, Sharova NI, Kuzmenok OI, Kotchergina NI, Nikonova MF, Litvina MM, Savina NP: The role of the thymus in the appearance of immunologic effects of low dose irradiation. Radiat Prot Dosim 62:77-79,1995

18 Kr. 100,-

J^^&i-t-t-iifir,.^1*"*^

ISSN 0804-4910 ***•:.;