Get documents about "Biological effects of low doses of ionizing radiation
Document Sample
SPECIAL REPORTS
Biological effects of low doses of
ionizing radiation: A fuller picture
The two latest reports of the United Nations Scientific Committee
on the Effects of Atomic Radiation (UNSCEAR)
provide a comprehensive overview of current knowledge
hen the United Nations Scientific Commit- diation doses also could induce serious health by Abel J.
tee on the Effects of Atomic Radiation (UN- effects, although of low incidence and only de- Gonzalez
SCEAR) submitted its 1994 report to the United tectable through sophisticated epidemiological
Nations General Assembly this year, the interna- studies of large populations. Because of the work
tional community received a fuller picture of the of UNSCEAR, these effects are now better and
biological effects of low doses of ionizing radia- more widely understood and better quantified.
tion. The 272-page 1994 report specifically ad- Effects at the cellular level: DNA damage
dresses epidemiological studies of radiation car- and repair mechanisms. The biological effects
cinogenesis and adaptive responses to radiation of radiation derive from the damage it causes to
in cells and organisms. the chemical structure of the cell. For low radia-
The report is designed to supplement the tion doses, damage to the deoxyribonudeic acid
more extensive 928-page report that UNSCEAR (DNA) in the cell's nucleus is of concern. The
presented to the UN in 1993.* That report ad- damage is expressed as DNA mutation occurring
dressed global levels of radiation as well as ma- in genes in chromosomes of stem cells, which
jor issues of radiation effects, including the can alter the information that passes from a cell
mechanisms of radiation oncogenesis; the influ- to its progeny.
ence of level of dose and dose rate on stochastic While DNA mutation is subject to efficient
effects of radiation; hereditary effects of radia- repair mechanisms, the repair is not error free.
tion; radiation effects on the developing human Most damage is repaired, but some damage re-
brain; and late deterministic effects in children. mains or is badly repaired, and this has conse-
Taken together, these two reports provide an quences for the cell and its progeny. (See box,
impressive account of current knowledge on the page 38.)
biological effects of ionizing radiation. This arti- Evidence of cell adaptation. There is experi-
cle — though by no means an account of all mental evidence that DNA mutations can be re-
essential information — summarizes the high- duced by a small prior conditioning dose of ra-
lights of UNSCEAR's assessment of the effects diation, probably because of stimulation of the
of low doses of ionizing radiation, hereinafter repair mechanisms in cells. (See box, page 42.)
called " low radiation doses" (see box, page 39) Such a process of adaptive response has been
in the context of available radiobiological evidence. demonstrated in human lymphocytes and in cer-
tain mouse cells. The cellular response is tran-
sient and there appear to be individual variations.
Radiobiological effects: As it is recognized that the effectiveness of DNA
The current understanding repair is not absolute, adaptation is likely to oc-
cur together with the processes of DNA mutation
Since the beginning of the 20th century, it has and its subsequent effects. The balance between
been known that high doses of ionizing radiation
produce clinically detectable harm in an exposed
individual that can be serious enough to be fatal. * See the 1994 UNSCEAR Report, Sources and Effects of
Some decades ago, it became clear that low ra- Ionizing Radiation; UN Pub. Sales No. E.94.IX.11; United
Nations, New York, (1994), and the 1993 UNSCEAR Report:
Sources and Effects of Ionizing Radiations; UN Pub. Sales
No.E.94.IX.2.; United Nations, New York (1993). Also see
Dr. Gonzalez is Deputy Director of the IAEA Division of the IAEA Bulletin. Vol. 35, No. 4, page 49 (1993), for high-
Nuclear Safety. lights of the 1993 report.
IAEA BULLETIN, 4/1994 37
SPECIAL REPORTS
Radiation Interaction of radiation with biological material
Exposure and affects the smallest unit of living matter capable of Ionizing
Living Matter independent existence: the cell, (a). A typical cell is Radiation
a sack of fluid, or cytoplasm, enclosed by a mem-
brane, which embeds a nucleus containing the chro-
(a)
mosomes — threads of complex biological sub-
stances, including the more essential compound of
life, deoxyribonucleic acid or DNA — that carry
life-sustaining information. The chromosomes hold
the genes, a segment of DNA that codes the informa-
tion, and allows its transmission from a cell to its
descendants. The cytoplasm also embeds organelles
governing important metabolic functions of the cells
and the generation of vital energy.
The human body contains a total of around one
hundred trillion (or 10 ) cells. They are variable in
shape and size, the average diameter being lower than
10 micrometres. The large majority of cells are so- (d)
matic cells, i.e. those which make up the bulk of the
organism. A relatively minor number of cells pass on
hereditary information from the organism to its de- (e)
scendants during reproduction: they are called germ
cells1. From the large number of human cells, only a
fraction has stem-like properties, i.e. are able to re-
produce a progeny of cells. The human body contains
a total of around 1010to 101 'of these stem cells; their
fraction varies among tissues and organs, and also
with age.
Radiation can ionize any atom in the cell compo- two chains—or strands—wound around each other.
nents. An important outcome is the production of This complex molecule comprises numerous individ-
active chemical radicals, extremely reactive com- ual units or nucleotides (d). Nucleotides are made
pounds, able to promote chemical changes in the cell. of four types of complementary bases called adenine
These changes may either damage essential cellular and guanine and thymine and cytosine. The se-
functions, and potentially kill the cell or prevent it quences of the bases express the genetic code.
from reproducing, or alter the genetic information. Directly, or indirectly by the action of chemical
The target cells for the radiation effects that are radicals, radiation can induce changes in the se-
expressed as a modification of the cell's genetic quence of bases and therefore alter the genetic code.
information are the stem cells. Interactions of radia- This process is referred to as mutation, or a sudden
tion with cell material may occur at random at any random change in the nucleotide sequence of a DNA
moment during the dynamic process of reproduction molecule (e), resulting in alterations in the genetic
of stem cells. At low radiation doses, there may be a code that, as a consequence, may cause the cells and
great deal of incident radiation per cell but the fre- all cells derived from it to differ in appearance or
quency of interactions is extremely low. UNSCEAR behaviour — referred to as a change in phenotype.
estimates that a low radiation dose (e.g. 1 mSv per Possible alterations are point mutation, or replace-
annum) will produce, on average, circa one interac- ment of one nucleotide by another, and clastogenic
tion per cell in a year. mutation including insertion or deletion, which is
The human cell contains 46 chromosomes (b) the addition or removal of any piece of DNA, from
and a large number of genes that determine the char- one base pair to quite extensive parts, and inversion,
acteristics of an individual. Genes exist in alternative which is the excision of a portion of the double helix
forms called atteles — one from each parent — followed by its reinsertion in the same position but
which occupy the same relative position in chromo- in reverse orientation. Mutation is passed on from an
somes having the same structural feature. One allele individual to his or her progeny during reproduction
may be dominant over the other, determining which via the germ cells.
aspect of a particular characteristic the organism will A cell or organism whose phenotype has been
display; the only "dominated" allele is known as altered by mutation is referred to as a mutant. The
recessive. more common generator of mutants is random errors
The gene component, DNA (c), is a pair of linear in DNA replication during cell reproduction. The
long chain-like molecules called polynucleotiaes mutation rate is increased if the cell is exposed to
wrapped around one another, as a spiral ladder- physical or chemical mutagens or agents able to
shaped double-helix complex molecule composed of cause mutation. Heat is probably the most important
38 IAEA BULLETIN, 4/1994
SPECIAL REPORTS
environmental mutagen. Radiation is a rather mild Radiation Doses
mutagen.
The term radiation means energy propagating in the form of electromag-
Mutation is effectively repaired by the cell
netic waves or photons, or in the form of subatomic particles. Ionizing
through mechanisms which are not yet well under-
radiation is radiation of sufficiently high energy to cause — in the medium
stood. It is likely that, if a point mutation occurs in
through which it passes — the production of pairs of ions, i.e. of atoms or
just one base of one DNA strand, repair would be
groups of atoms that have either lost or gained one or more electrons to
easy as the complementary base in the other strand
become positively or negatively charged, and the corresponding complemen-
apparently can act as a template for the repair, but for
tary electrons. For biological effects, the medium in which ion pairs are
mutations occurring in the same location of both
produced is biological material, more specifically cellular material.
strands, or if clastogenic damage occurs, error-free
The term radiation (absorbed) dose generally means the amount of
repair would be less likely .Radiation seems to be a
energy which is absorbed from ionizing radiation by a unit mass of material.
stimulant of the repair process. (See box on Adaptive
This quantity is expressed in unit energy per unit mass, that is in joules per
Response, page 42.) It seems, however, that there is
kilogram, which takes the special name gray (Gy); [1 Gy = 1000 milligray
always a chance of misrepair, even in single strand
(mGy)]. For radiation protection purposes, the absorbed dose is weighted to
point mutations.
take account of the effectiveness of different radiation types and the radiosen-
Unrepaired mutation is responsible for the detri-
sitivity of various organs and tissues. The resulting quantity is termed
mental fate of a mutated cell. If a mutation is not
effective dose, and its unit sievert (Sv) [ISv = 1000 millisievert (mSv)]; for
properly repaired, the outcome for the cell can be
photons in the intermediate energy range, 1 mGy is approximately equal to
twofold: either the cell dies — for instance through
1 mSv.
apoptosis — or it survives as a viable but trans-
The term low radiation dose is used to mean a radiation dose lower than
formed cell that may give rise to a new family of
designated levels; sometimes it is also informally used to mean a low dose
mutant cells. The two outcomes will have very dif-
rate, i.e. low dose per unit time. In specialized radiobiological forums, low
ferent consequences for the organism. At low radia-
radiation dose (and dose rate) refers to exposures for which it is very unlikely
tion doses, the killing of cells is sparse and does not
that more than one event of energy absorption from radiation will occur in
usually have serious health consequences. But a mu-
the critical parts of a cell (and damage it) within the time during which repair
tant cell can evolve to cause serious health effects: if
mechanisms in the cell can operate. Thus, UNSCEAR concluded that low
it is a somatic cell, it can be the initiator of a malig-
radiation dose refers to a total dose of less than 200 mSv and dose rates below
nancy, and if it is a germ cell, of hereditary diseases.
, 0.1 mSv per minute (which in fact is a very high dose rate of around 5000
mSv per annum).
For the non-specialized public, low radiation doses are deemed to
correspond to levels similar to those from, for instance, natural background
exposure or some very common radiation exposures such as those arising
during air travel. Natural background exposure varies widely around the
world. Some "normal" [and "elevated"] values of annual dose rates are as
follows: for cosmic rays, 0.38 mSv [2.0 mSv]; for terrestrial radiation 0.43
mSv [4.3 mSv]; and for exposure to radon, 1.2 mSv [10 mSv]; leading to an
average total of around 2.4 mSv per annum. The average annual dose for
very frequent flyers (such as aircrew) is around 2.5 mSv. These dose rate
Germ cells are: the testis seminiferous tubule cells which levels of a few mSv per annum are expected to deliver, during a lifetime,
divide by mitosis into spermato gonia and then into sper- doses of above around 100 mSv, which are of the order of magnitude of the
matocytes, followed by a meiosis into spermatids which low radiation doses designated by UNSCEAR.
eventually develop into spermatozoa; as well as the special
oogonia cells within the ovary which divide by mitosis into
oocytes which after two meiotic divisions become an ovum.
The fusion of a spermatozoon and an ovum forms a zygote,
the origin of a new being. stimulated cellular repair and residual damage is
2
not yet clear.
Apoptosis is an orderly, systematic and programmed
Dose-response relationship. If DNA muta-
process of self-destructive death of the cell. Probably as a
result of genetic altrations, the cell enters into a period of tion depends on radiation's interaction with a single
cytoplasmic basophilia and nuclear condensation, followed cell, then the frequency of DNA mutation — in
by eosinophilia and cytoplasmic condensation, cell frag- cases of no interaction between cells — should
mentation, and dissolution and, typically, phagocytosis by
follow a linear-quadratic relationship with dose.
neighbouring cells. Contrary to cell terminal differentia-
tion, which is a cessation of cell replication, to cellular (See box, page 42.) Furthermore, if it is assumed
senescence, which becomes manifest only at the end of die that, for low radiation doses, mainly single interac-
life span of the cell, and to the disorganized cellular death tions of radiation rather than multitrack effects are
by necrosis, apoptosis is an orderly cellular process of
dominant, the frequency of cells with one or more
self-destruction which can be initiated at any moment in the
cell life. It is speculated that radiation can be an important interactions, and consequently the frequency of
initiator of apoptosis which might have a potentially bene- DNA mutations, will simply be proportional to
ficial influence in tumour promotion and malignant dose. Thus, if a fraction of mutations remain unre-
progression.
paired, the expected number of mutated cells will
be proportional to the dose.
IAEA BULLETIN, 4/1994 39
SPECIAL REPORTS
Cell killing: deterministic effects. A number may not play a major role in moderating human
of radiation interactions in the cell and some of radiation carcinogenesis. However, specialized
the unrepaired DNA mutations may lead to the immune functions in certain organs and the exist-
death of the mutated cell, or prevent it from ence of non-immunogenic cell surveillance mecha-
producing progeny. This may occur as a result of nisms suggest that a proportion of early pre-neo-
the cell's necrosis (i.e. its pathological death as plastic cells may be eliminated before they become
a result of irreversible radiation damage) or established. Other mechanisms defending against
apoptosis (i.e. a programmed self-destruction of tumour induction and development include the al-
the cell) or because the normal cellular reproduc- ready mentioned DNA repair, apoptosis, terminal
tion is hindered. For low radiation doses, cell differentiation and phenotypic suppression. Alto-
killing is sparse and therefore of no negative gether, these mechanisms will reduce the prob-
consequence to health owing to redundancy of ability that a specifically damaged target cell will
cellular functions and cellular replacement. For progress to frank malignancy; to estimate this
high radiation doses which could kill large num- probability, however, is extremely difficult.
bers of cells in an organ or tissue, the cell-killing Adaptive response in organisms. Evidence
effect could be lethal for the tissue and, if vital of organic adaptive response to radiation expo-
tissues are involved, for the individual con- sure in laboratory mammals has been reported in
cerned. Although killing of individual cells oc- the literature. However, because of the lack of
curs at random, the health effects resulting from conclusive evidence, UNSCEAR remains doubt-
the extensive cell killing at high doses are called ful whether adaptation also occurs at the cellular
"deterministic effects" because they are prede- system level and whether the immune system
termined to occur above a threshold level of plays any role in the process.
dose. Deterministic effects, therefore, are not Epidemiological evidence of carcinogene-
clinically expressed at low radiation doses. Ex- sis. Although it is not yet possible to determine
ceptionally, the killing of a few essential cells clinically whether a specific malignancy was
during organ development in utero may result in caused by radiation, radiation-induced tumours and
severe harmful effects clinically expressed in the leukaemia have been detected and statistically
newly born; these effects are generally referred quantified by epidemiological studies of popula-
to as "effects in embryo". tions exposed to relatively high radiation doses.
Cell transformation: stochastic effects. From initiation until the clinical expression of the
Other unrepaired DNA mutations may produce cancer, a period of time — termed the latency
modified but viable stem cells. If the modified period — elapses. The duration of the latency
cell is a somatic cell, it can be the initiator of a period varies with the type of cancer from a few
long and complex process that may result in years in the case of leukaemia to decades in the
severe " somatic health effects", such as cancer. case of solid tumours. The action of radiation is
Alternatively, if the cell is a germ cell, the muta- only one of many processes influencing the de-
tion could be expressed as hereditary health ef- velopment of malignancies and, therefore, the
fects in the progeny of the exposed person. These age at which a radiation-induced malignancy is
health effects, both somatic and hereditary, de- expressed has been found to be no different from
riving from a cell modification are called "sto- the age for malignancies arising spontaneously.
chastic effects" because their expression is of an Epidemiological studies of a number of
aleatory, random nature. populations exposed to generally high-dose and
high-dose-rate radiation — including the survi-
vors of the atomic bombing of Hiroshima and
Carcinogenesis Nagasaki in Japan and patients exposed in thera-
peutic medical procedures — have provided un-
A most important stochastic effect of irradia- equivocal association between radiation dose
tion is carcinogenesis. It is believed to be a multi- and carcinogenesis.
stage process and is usually divided, albeit im- The most comprehensive source of primary
precisely, into three phases: cancer initiation, tu- epidemiological information is the Japanese
mour promotion, and malignant progression. (See survivors' "life span study". This has demon-
box, page 41.) It is presumed that radiation is strated a positive correlation between the radia-
important as an initiator rather than as a promoter tion dose incurred and a subsequent increase in
or progressor. For low radiation doses, therefore, the incidence of, and mortality due to, tumours
as the likelihood of initiating mutations is pro- of the lung, stomach, colon, liver, breast, ovary,
portional to dose, the likelihood of carcinogene- and bladder, and also of several forms of leukae-
sis should also be proportional to dose. mia but not for lymphoma or multiple myeloma.
Immune response and cell surveillance Of the 86 300 or so individuals in the " life span
mechanisms. It is argued that immune response study" cohort, there were 6900 deaths due to
40 IAEA BULLETIN, 4/1994
SPECIAL REPORTS
Carcinogenesis: A Multistage Process
Carcinogenesis is believed to be a multistage process usu- facilitated by additional alterations in initiated and promoted
ally divided into three phases: cancer initiation, tumour promo- cells to become promoter - independent and invasive. The
tion, and malignant progression. principal phenotypic characteristic of the malignant progression
Cancer initiation. Most, if not all, cancers seem to " initiate" is the ability to spread, or metastasize, from the primary tumour
from DNA mutation in a single stem cell which thus becomes a mass and to establish secondary growth foci, or metastases, at
modified, carcinogenic cell. This process involves loss of control other sites. This is a complex, multifaceted process that appears
over the cellular reproduction cycle and differentiation. It is pre- to involve a series of subsequent genetic changes within the
sumed to start as a result of deactivation of tumour suppressor evolving pre-neoplastic clone of cells, including changes in
genes that seem to play a crucial role in regulating cellular prolif- growth rate, growth factor response, invasiveness, and metas-
eration. The loss in activity of these genes, through for instance a tatic potential. The progression stage includes angiogenesis,
deletion or a mutation, can lead to uncontrolled cell growth. The detachment, invasion, release, survival (host interaction), arrest,
process of initiation of Carcinogenesis might also be the result of extravasation and invasion, new growth, angiogenesis. Thus the
conversion of proto-oncogenes, which seem to be involved in process is repeated until clinically important metastases are
regulating the proliferation and differentiation of cells and can produced. Whether and how radiation exposure influences the
potentially become oncogenes and transform the cell into a malig- changes leading to progression and the different stages of the
nant cell. Relative target sizes for the induction of these events progression process is not yet known. The progression stage also
would tend to indicate tumour suppressor genes as the most appears to be irreversible.
radiosensitive targets. It is presumed that the initiating event
centres on single gene deactivation in a number of possible
genes and that initiation is an irreversible process.
Tumour promotion. The promotion stage involves the clonal Death necrosis, apoptosis
expansion of an initiated stem cell into a focus of non-terminally
differentiated cells. The initiated cell can be stimulated or "pro-
moted" to reproduce by agents that, alone, may have low carcino-
genic potential but that are able to enhance greatly the yield of
neoplasms induced by prior exposure to an initiator. Radiation, like
many other agents, can act independently as initiator and promoter.
After initiation, the transformed cell may have some proliferative or
selective advantage over normal cells, such as a shorter reproduction
time. However, the transformed cells and their immediate progeny
are surrounded by normal cells, which constrains their pre-neoplastic
properties as they are prone to be eliminated in the competitive
reproductive process. Elimination becomes more unlikely as the Self-renewal
number of transformed cells increases. Thus, the promotion stage
seems to be potentially interruptible and reversible.
Malignant progression. After initiation and promotion, a
futher stage of " progression" is needed to complete the multi- I Cancer initiation % Tumour promotion • Malignant progression
stage Carcinogenesis. It is characterized by a progressive ten-
dency towards increasing malignancy. Progression might be
solid tumours during 1950-1987, but only ap- Epidemiological studies on the effects of low-
proximately 300 of these cancer deaths can be dose-rate exposure undertaken for occupational ex-
a t t r i b u t e d to r a d i a t i o n exposure. The posures have shown conflicting evidence. While a
epidemiological data for leukaemia incidence in number of occupational studies have reported a sig-
this same period indicate statistically that 75 nificant excess risk of leukaemia in workers exposed
cases out of a total of 230 leukaemia deaths can to radiation — which is broadly in agreement with
be attributed to radiation exposure. The inci- the estimates derived from high-dose-rate studies
dence data also provide evidence of excess for — other studies have failed to demonstrate any
thyroid and non-melanoma skin cancers. The positive correlation. (See author s note, page 45.)
study provides little or no evidence of radiation Studies of lung cancer in miners occupationally ex-
induction for cancers of the rectum, cervix, gall posed to radon, however, have been able to provide
bladder, larynx, prostate, uterine cervix, uterine a consistently positive correlation between excess
corpus, pancreas, kidney, renal pelvis, or testes, cancer incidence and radiation dose.
or for chronic lymphocytic leukaemia and Many environmental exposure studies have
Hodgkin's disease. been carried out, notably on the incidence of
IAEA BULLETIN, 4/1994 41
SPECIAL REPORTS
Adaptive Response
The possibility has been known for many years that low doses include gene coding for transcription factors — i.e. factors affect-
of radiation may cause changes in cells and organisms, reflecting ing the process of transfer of genetic information of DNA — and
an ability to compensate for the effects of radiation. It has been synthesis of enzymes involved in the control of the cell cycle and
suggested that estimates of the risk of stochastic effects from therefore in the proliferation of cells as well as in the repair of
low-level radiation may have been overstated because no allow- damage. Observations support the hypothesis that conditioning
ance has been made for this process, which is referred to as doses activate certain genes and that this is quickly followed by
adaptation or adaptive response. The term adaptive response is the synthesis of enzymes responsible for DNA repair. If these
used to refer to the possibility that a small dose of radiation — enzymes become available in adequate concentration at the time
which is variously referred to as the adapting, inducing, priming, the cells receive the challenge dose, the extent of DNA repair
or conditioning dose — may condition cells by inducing proc- seems to be improved. The adaptive response mechanisms are
esses that reduce either the natural incidence of malignancies or thought to be similar to those operating after exposure to other
the likelihood of excess malignancies being caused by a further toxic agents, including trace amounts of oxidizing radicals. The
radiation dose—usually referred to as the challenge dose. In vitro adaptive response to radiation, therefore, may be the result of a
adaptive response of lymphocytes takes place between about four general mechanism of cellular response to damage.
and six hours after exposure to a conditioning dose within a range
of dose of about 5 mGy to 200 mGy, and remains effective for
around three cell cycles. Following a challenge dose, repair is
manifested as a reduction — below the expected levels — in Conditioning dose Response
chromosomal aberrations, sister chromosomatic exchanges, in-
duced micronuclei, and specific locus mutations, sometimes by a
factor of about two. Moreover, bone marrow cells and spermato-
cytes from mice exposed to a challenge dose that followed a
conditioning dose also showed reduction in the number of chro- Challenge dose
t
Response
mosomatic breaks compared with cells exposed to the challenge
dose alone.
It seems that many agents can be activated sometime after n
Conditioning dose Challenge dose
I
Response
exposure to the conditioning dose and can reduce DNA mutations
due to the subsequent exposures to the challenging dose. These
Dose-Response Relationship
It is presumed that radiation acts through single track interac- the available epidemiological data suggest that for tumour induc-
tions occurring randomly according to a Poisson distribution in a tion the DDREF adopted should have a low value, probably of
homogeneous population of cells. It can be mathematically shown around two and no more than three. In the case of hereditary
that a linear-quadratic expression describes the theoretical dose- disease, a DDREF of three is supported by most experimental data
response relationship — i.e. the mathematical relation between for animals.
the dose incurred and the probability of expression of an attribut-
able radiation effect. This relationship fits most of the available
epidemiological data. For low radiation doses, there are so few Incidence of cancer
(induced by radiation exposure)
radiation tracks that a single cell (or nucleus) is very unlikely to
Linear-quadratic
be traversed by more than one track. Thus, under these assump- relationship
tions the dose-response relationship is almost bound to be linear,
independent of dose rate and without dose threshold.
Since most available radioepidemiological data are for high
doses only, the approach commonly used for assessing the risk at Ideally fitted
linear
low doses is to fit an ideal linear dose-response relationship to the relationship
data in order to project it for low doses for which data are lacking.
As the real dose-response curve is assumed to follow a linear-
quadratic relationship, with the linear term prevailing at low
doses, a reduction factor — which is called "dose and dose rate Linear term
effectiveness factor" or DDREF — has to be applied . Based on
experimental data, it seems that the DDREF should be small. For
cell tranformation and mutagenesis in somatic and germ cells,
per 1000 mSv Dose
DDREFs of around two or three have been observed, although at
low dose rates no reduction in effects was observed (i.e. the • Epidemiological data
DDREF was unity), over a large range of doses. Taken together,
42 IAEA BULLETIN, 4/1994
'Niianna vavi
si asop Suijqnop sqj 'snqjj 'uoijBindod sqj in aDuai aqj uo KjuiBjaaDun aiqBjapisuoD qjiA\ pajBiuijsa
-jnDDO [BJ3U3§ [BjnjBU SJI OJ UOIJE|3J UI UOIJBIpBJ si qaiqA\ 'OMJ jnoqB jo JOJDBJ uoijDnpaj y 'B1BP
A"q paonpin Suiaq assssip AJEjipsjsq B jo pooqip^i [BDi§ojoiiuapida ajBJ-asop-qSiq pire asop-qSiq
sqj Suissajdxa JE suire jj -auiBJij aausjsjsj B SB (SSBS aqj oj jij (pioqsajqj-uou) JBauij jBDijajoaqj B IUGJJ
-sip B ipns jo) SDUsjBASjd [BJTIJBU sqj §uisn 'uotj paanpap >jsu aqj oj pai|ddB aq pjnoqs JOJDBJ uoij
-BipBJ oj psjnqujjB ssBssip AJBjipajaq jo sasBD jo -Dnpaj y 'sajBiuijsa >(su aoj pasn aq jsniu sasop
jaquinu [BuoijippE aqj jo sum) ui ajBiuijsa aqj sapiA qSiq JB BjBp aauappui jBqj sireaui sasop /v\o( JB
-aid po\\i3\u (uoiwjnw 3AiiDj3j JQ) ssop Sui/qnop BiiuaB>(na| PUB jaDUBD jo uoijDnpui aqj uo BjBp JBD
aqj_ -sjuauiuadxa reuiiuB uiojj BjEp ssn qDiqM spo -i§o|oiiuapid3 jo >(DB[ aqj 'pireq jaqjo aqj UQ
-qjaui jDSJipui OAYJ qSnoiqj psjBunjss are sireuinq •(sjapoui aaqjo aqj qjiM jaqSiq aq UBD asBD paonp
in SJDSJJS AJBjipajaq paanpin uoijBipBJ jo ssDusp -ui aad jsoj sjBa/( aqj anqM japoui JUBJSUOD aqj
-pin 'SDUSPIAS [EDi8o|oiuiapida jaanp jo 5jDBj aqj qjiA\ jaq§iq si qjBap paanpui ajnsodxa jo >[su aqj)
JO A\3IA U] 'SjapJOStp &JDJlpdl3tf JO/ S]3pOffl ajnsodxa aqj jajjB aiuij aiuos asBaJDap IJIM ji jBqj
•ppqsajqj asop ou qjiM 'asop qjiM JBSUII gq sjaqjo aqj 'aji| jnoq§nojqj JUBJSUOD SUIEIUSJ ajBJ
oj pauinsaad OSJB si sasop uoijBipBJ MOJ JB asuods aAijBjaj ssaaxa aqj jBqj saiunssB auo :aV33SNfl
-ai aqj 'ajojaaaqj_ -JOJBIJIUI UB Suiaq jo ajqBdBD Xq pasn ajB suoijDafojd 3AijBDi|dij[nui aajqx
Xjjnj si uoijoBaajui uoijBipBJ XUB qjiM japjosip '(3§B qjiM S3SB3JD
aqj jo uiSuo jjao a|§uis oijsBqoojs :JB[IUIIS are -ui OSJB qDiqM) 3JBJ JSDUBD snoauBjuods 3qj
apBiii suoijdiunssB aqj jnq sisauaSoupJBD jo jeqj oj uoijjodojd ui 'a§B qjiM SSBSJDUI [[IM SJSDUBD
ireqj poojsjapun jjaM ssa| si uoijBipBJ uiojj saapjo p3Dnpui jo SJBJ sqj jBqj ssuinssB qDiqM '(\3poiu
-sip AJBjipaaaq jo uoijBjauaS jo ssaaoad aqx it3Aiuppv,, ajdiuis B qSnojqj UBqj jaqjBJ) \apoiu
•SJOJDBJ jBjuamuojiAua pire oijauaS jBaa t3M}D3}]dn]mu„ B qSnojqj auop MOU si siqj_
-A3S JO UOIJ3BJ3JUI 3qj UIOJJ guIJ(nS3J SUOIJBjniU •uoijBindod aqj jo auiijajn aJijua aqj J3AO UOIJBA
oj snp sjapjosip iDuojjv/jjjnw 'XijBijusjod 9 -jasqo jo pouad aqj Suunp pajou SJSDUBD ssaaxa
puB tuoijBjndod B ui jusssad aiB jBqj sauaS aqj jo uoijDnpui jo ^Duanbajj aqj jDafojd oj AJBS
jo ajoq/w aqj u; -3-1 '(ood susS s^uoijBindod sqj ui -saDau si ji 'uoijBjndod pasodxs UB ui >|sij siuij
ajBinuinooB ^BUI jnq suoijEasusS Maj jsjy aqj ui -ajij jjnj aqj uiBjqo oj japjo ui 'aDUBjsui JGJ -sja
jDajjs 3iJji[ ssonpoad qoiqM puB — jiBASjd p|noM -poui iBDiSojoiqoipBj a[qisnBjd uiojj suoijdiuns
ajajjB jireuiuiop aqj 'asi/vuaqjo — sjuajBd qjoq -SB (BDijajoaqj jo JUHODDB SUIJJBJ 'ajqB[iBAB BJBP
uiojj pajuaqui aq X[uo UBD qDiqM '3\3\]v 3MSS33 jBDiioioiiuapida pajiiuq aqj UIDJJ uoijBjodBJj
-djt aqj ui uoijBjnui B -a'l) uoijBjnui aAissaoaa « -xa Xq jno paujBD ajB sisauaSoupJBD jo sjuaiu
;(suoijBJauaS jBjaAas qSnojqj passaadxaun -ssassB 5jsi^ -sissuySouiJJDJ jof sppof^
passBd aq UBD pire uoijBaauaS jsjij aqj ui sjap •SaSOp MOJ JB UOlJBIpBJ JO SJD3JJ3 3qj OJ SuiMO
-josip oj spBaj qoiqM pire juaiBd auo Xjuo UIGJJ suBuinq ui JSDUBD jo s^jsu SSSBSJDUI jo SSSESJD
pajuaqui aq UBD qoiqM 'auaS B jo 3]3{]v iimwiuop -sp jsqjp s[jaD ui asuodssj aAijdBpB aqj JBqj
aqj ui uoijBjniu B 'a-i) uoijBjniu jireuiiuop 0 M3iA sqj yoddns oj 3DuspiAa aAisnpuoD ou apiA
:jo jjnsaj aqj aq ABUI sjoajja -ojd BjBp ireiunq pajiuiq pire sjusiuusdxs JBIUIUB
AjBjipajaq [Bijuajod aqj, 'ajnsodxa UOIJBIPBJ Sui 3AISU3JX3 3qj 'J3AO3JO^ 'SJ3DUBD UBUinq
-MO[joj sireiunq ui sjoajja AjBjipaaaq jo uoijonpui -DO X[snoairejuods jo aausjBASjd sqj ui
aq paapui UBD ajaqj jBqj pauinssB XjaAijBAjas B SB p3SS3jdX3 3SUOdS3J 3AIjdBpB UB JO 3DU3p
-UOD si ji 's§uipuy UBiunq 3AijB§au aqj jo SUOIJBJ -IAS spiAOjd jou op Xaqj 'ajojajaqx -JaA\od |BD
-iiuq (BDIJSIJBJS aqj jo JUPODDB Sui5(Bj puB 'saipnjs -IJSIJBJS J3MO| jo uaaq aABq uoijBjdBpB uo saipnjs
jBjnuaD pire siusiireSjo jo aSirea apiM B qjiM uotj |BDi8o|oiuiapida ireuinq aqx -asuodsyj anydvpn
-Bjuaiuuadxa DijauaS jo sissq aqj uo 'jaA3A\oj-i fo 93uapua ]D3i8ojoiiuapida aAisnjyuoouj
•suBiunq ui uoijBipBJ jo sjaajja A^rejipsjsq pajaaj •SUOIJBIDOSSB JUBDIJIuSlS X[IBDIJSIJBJS
-ap 'aDuapijuoD jo aajSap JUBDIJIU§IS XJJBDIJSIJBJS Xire paanpojd jou aABq uoijBipBJ punoj§^DBq JBJ
B qji/v\ 'jou aABq saipnjs [BDiSopiiuapidg -[Bn -njBU oj ajnsodxa jo sjaAaj MO) puB qSiq jo sBajB
-puipui pasodxa aqj jo sjirepuaDsap aqj ui sjap UI 3DUappUI J3DUBD JO SUOSUBdllKQ "SS3DXa UI
-uosip LiDi}p3i9\\ SB jsajiireui auioDaq pire suon aq oj punoj SB/V\ BiiuaB^naj uioqM SUOIUB '
-BJaua§ juanbasqns oj pajjiiusirejj aq ajdpuud jaiujoj aqj ui jaAiy BqDax aq) ojui
ui pjnoD [PD aqj JQJ jBqjaj-uou 3JB jBqj sjjao aAijDBOipBJ jo sa§JBqDsip pAaj qgiq oj pasod
jBuiuua§ ui suoijBjniu VNQ pa-iredajun /(uy -xs 3]doad uo Xpnjs B si uoijdaDxs jBjnDijjBd y
•ajnsodxa UOIJBIPBJ oj pajnqujjB aq UBD sjajsnp
qDns jBqj Xp^ijun si ji jBqj sajBDipui aauapiAa
jaqynj 'SUOIJBJIBJSUI JBa]Dnu jo Xjiiuixojd aqj puB
sasBD BiiuaE5jna[ jo sjajsnp uaaMjaq
Clp aSod 'xoq 39$) -sjuauissassB 5jsu sji -JOD aAijisod papiAOjd aABq oj pauodaj
u aj3M saipnjs qDns MSJ B qSnoqjjy 'S
! HV3DSNfl ^1 P3sn s! 'BJBP IBDiSojoiuiapida
aiuos puB suoijdiunssB jBDtjaJoaqj jo sisBq -ui JB3pnu JBau SUIAIJ suoijBjndod ui BiiuaB>(n3[
sibOd3b
SPECIAL REPORTS
the dose expected to produce as many mutations mates in relation to health effects of low radia-
as those that occur spontaneously in a generation tion doses. As a result, the scientific body contin-
and it is obtained by dividing the spontaneous ues to consider that radiation is a weak carcino-
mutation rate in a locus — or position — of a gen and an even weaker potential cause of he-
relevant gene in a chromosome by the expected reditary diseases. A summary of UNSCEAR's
rate of induction of mutations per unit dose.) The quantitative estimates follows:
direct (or absolute mutation) method directly as- • Epidemiological Estimates:
sesses the expected incidence of hereditary diseases Lifetime mortality:
by combining the number of genes at which muta- D 1.1% after exposure of 1000 mSv for
tions can occur with the expected number of muta- leukaemia and 10.9% for solid tumours
tions per unit dose and the dose itself. It is therefore (12% in total). For reference, in UN-
aimed at expressing the likelihood of hereditary SCEAR's 1988 report, the correspond-
diseases absolutely, in terms of the expected in- ing data was 1.0% for leukaemia and
crease in the prevalence of the disease. The esti- 9.7% for solid tumours.
mates of risk do not usually include the many n linear between 4000 mSv and 200 mSv
hereditary diseases and disorders of complex, mul- (little evidence at lower dose).
tifactorial aetiology, in view of the fact that any
effect of radiation upon the incidence of multifac- • Radiobiological Estimates:
torial disorders should be only slight and is highly For low (chronic) radiation doses of around
speculative. 1 mSv per year:
D probability of an excess malignancy:
10"4 per year
Effects on the embryo n lifetime probability: 0.5%
D proportion of fatal concerns in the popu-
Effects of radiation in utero are generally lation that may be attributed to radia-
referred to as effects on the embryo. They can tion: approximately 1 in 40.
occur at all stages of embryonic development, The above estimates are based on the follow-
from zygote to foetus and may include lethal ing assumptions and inferences:
effects, malformations, mental retardation and Assumptions: ;;
cancer induction. The first three may be the D cells in the human body: 101 cells per
possible outcome of deterministic effects during individual
embryonic development, particularly at the pe- n target stem cells: 1010 to 1011 cells per
riod of formation of organs. individual
Evidence of effects on brain growth and de- n initiating event: single gene mutations in
velopment has emerged after observations of se- one of around ten possible genes
vere mental retardation in some children exposed D induced mutation rate (per cell): 10" per
in utero at Hiroshima and Nagasaki. The effects 1000 mSv
from high-dose, high-dose-rate exposure in D excess probability of malignancy:
utero, particularly linked to the period between 8 and approximately 10%; and
15 weeks after conception, seem to indicate a D interactions per cell: 1000 per 1000 mSv.
downward shift in the intelligence quotient (IQ) Inferences:
distribution. For low radiation doses, this poten- D excess malignancy: 1 per 1011 to 1012 target
tial effect on the embryo is undetectable in the cells receiving 1000 mSv;
newborn. D rate of target gene deactivation: 10" per
Studies of in utero exposures have given cell per mSv; and
conflicting evidence of carcinogenesis in the D probability that a single track will give rise
child, from relatively high risk to essentially to an excess malignancy: 10" to 10" .
small undetectable risk, including (possibly)
none at all. There is no biological reason to • Risk Estimates:
assume that the embryo is resistant to carcino- Risk of malignancies:
genesis but on the basis of current data such a lifetime probability of radiation induced
effects cannot be quantified with any certainty. fatal cancers:
5% per 1000 mSv in a nominal population
of all ages; and
Highlights of UNSCEAR's conclusions 4% per 1000 mSv in a working population.
Risk of hereditary effects:
Taking account of the available radiobiologi- (via doubling dose method)
cal and radioepidemiological information, UN- D probability of hereditary radiation
SCEAR has made a number of quantitative esti- effects for all generations:
44 IAEA BULLETIN, 4/1994
SPECIAL REPORTS
1.2% per 1000 mSv (or 1.2% per genera- a restricted capability to detect and quantify sta-
tion for a continued exposure of 1000 tistically significant stochastic radiation effects
mSv per generation) — both somatic and hereditary. As a result, un-
D probability of hereditary effects in the equivocal direct observational evidence of the
first two generations: effects of low level radiation does not exist and
0.3% per 1000 mSv will probably not be obtainable for a long time.
(via the direct method) Obtaining unequivocal evidence would require
D probability of hereditary effects (clini- sound epidemiological studies, able to associate
cally important disorders) in the first an increased incidence of specific health effects
generation: with radiation exposure. Such studies would
0.2% and 4% per 1000 mSv. have to overcome inherent statistical and demo-
Risk of effects on embryo: graphical limitations and moreover should in-
(for those exposed in utero in the period bet- clude correct case ascertainment, appropriate
ween 8 and 15 weeks after conception) comparison groups, sufficient follow-up, control
D downward shift oflQ distribution: of confounding factors and well-characterized
30 IQ points for 1000 mSv dosimetry. It is not now feasible to obtain such
D dose required to shift from normal IQ to evidence for the effects of low radiation doses
severely mentally retarded: and therefore a continuing lack of direct evi-
lOOOmSv or more dence on such health effects is to be expected.**
D dose required to shift from low IQ to Because of these limitations, radiation risk
severely mentally retarded: estimates have to rely on an idealized radiobi-
a few hundred mSv. ological model, intended to provide the basis for
interpreting the available epidemiological results
Taking UNSCEAR's estimates together and for high radiation doses. Although the model
adding to them an estimated detriment from non- reflects sound understanding so far, it is rather
fatal cancers, the International Commission on simple, perhaps even simplistic, and it is still
Radiological Protection (ICRP) has recom- evolving. Scientific developments are taking
mended the use — for radiation protection pur- place that will extend knowledge of the biologi-
poses — of total nominal risks from stochastic cal effects of radiation and may necessitate
effects of radiation of: changing the model. Research in molecular biol-
• 0.0073% per mSv for the whole population; ogy, for instance, may provide new information
and on the mechanisms of cancer induction. The
• 0.0056% per mSv for all adult workers. mechanisms of adaptive response and the role of
These have been the nominal risk factors used radiation exposure in the initiation, promotion,
in developing the new International Basic Safety and progression of cancer will be better under-
Standards for Protection against Ionizing Radia- stood. The coming years might change our view
tion and for the Safety of Radiation Sources.* of the health effects of low radiation doses.
Notwithstanding the rapid progress in rele-
vant scientific branches, UNSCEAR has not yet
Outlook found it necessary to make any major revision to
its perception of the biological effects of radia-
Thanks to the work of a unique body in the UN tion and the consequent risk estimates. Nearly a
system, UNSCEAR, the biological effects of ioniz- quarter of the human population incurs fatal ma-
ing radiation are better known than those of many lignancies but, as UNSCEAR indicates, only
other chemical and physical agents affecting human " about 4% of deaths due to cancer can be attrib-
beings and the environment. However, there are still uted to ionizing radiation, most of which comes
many unanswered questions in radiobiology, in from natural sources that are not susceptible to
particular in relation tp the effects of low radiation control by man". O
doses. One problem is the lack of empirical evi-
dence. It should be emphasized that at low dose
levels, epidemiological studies presently have only
** Author's note: At the time this article is being issued, the
* The standards were developed under the auspices of the IAEA International Agency for Research on Cancer is releasing the
and five other organizations: the Food and Agriculture Organi- results of an epidemiological study on cancer risk among
zation. International Labour Organization, Nuclear Energy 95 673 nuclear industry workers. The study gives the most
Agency of the Organization for Economic Co-operation and precise direct estimates of mortality due to protracted low
Development, Pan-American Health Organization, and World radiation doses. As reported in Lancet (344: 1039-43), the
Health Organization. For a report on the new standards, see the estimates "provide little evidence that the [UNSCEAR] es-
author's article in the IAEA Bulletin, Vol. 36, No. 2 (1994). timates...are appreciably in error".
IAEA BULLETIN, 4/1994 45
Related docs
