Download Radiation carcinogenesis: lessons from Chernobyl

Oncogene (2009) 27, S9–S18
& 2009 Macmillan Publishers Limited All rights reserved 0950-9232/09 $32.00
www.nature.com/onc
REVIEW
Radiation carcinogenesis: lessons from Chernobyl
D Williams
Christ’s College, University of Cambridge, Cambridge, UK
Radiation is a carcinogen, interacting with DNA to
produce a range of mutations. Irradiated cells also show
genomic instability, as do adjacent non-irradiated cells
(the bystander effect); the importance to carcinogenesis
remains to be established. Current knowledge of radiation
effects is largely dependent on evidence from exposure to
atomic bomb whole body radiation, leading to increases in
a wide range of malignancies. In contrast, millions of
people were exposed to radioactive isotopes in the fallout
from the Chernobyl accident, within the first 20 years
there was a large increase in thyroid carcinoma incidence
and a possible radiation-related increase in breast cancer,
but as yet there is no general increase in malignancies. The
increase in thyroid carcinoma, attributable to the very
large amounts of iodine 131 released, was first noticed in
children with a strong relationship between young age at
exposure and risk of developing papillary thyroid
carcinoma (PTC). The extent of the increase, the reasons
for the relationship to age at exposure, the reduction in
attributable fraction with increasing latency and the role
of environmental factors are discussed. The large number
of radiation-induced PTCs has allowed new observations.
The subtype and molecular findings change with latency;
most early cases were solid PTCs with RET–PTC3
rearrangements, later cases were classical PTCs with
RET–PTC1 rearrangements. Small numbers of many
other RET rearrangements have occurred in ‘Chernobyl’
PTCs, and also rearrangement of BRAF. Five of the
N-terminal genes found in papillary carcinoma rearrangements are also involved in rearrangements in hematological malignancies; three are putative tumor suppressor
genes, and two are further genes fused to RET in PTCs.
Radiation causes double-strand breaks; the rearrangements common in these radiation-induced tumors reflect
their etiology. It is suggested that oncogenic rearrangements may commonly involve both a tumor-suppressor
gene (or a DNA repair gene) as well as an oncogene.
Involvement of two relevant genes would give a greater
chance of progression and a shorter latency than a singlegene mutation. More information is needed on germline
mutations conferring susceptibility to radiation-induced
PTCs, particularly DNA repair genes. The radiation
exposure to the fallout after Chernobyl was very different
from the whole body radiation after the atomic bombs.
The type and molecular pathology of the thyroid tumors
Correspondence: Dr D Williams, Strangeways Research Laboratory,
Worts Causeway, Cambridge CB1 8RN, UK.
E-mail: [email protected]
is changing with increasing latency, long latency tumors in
other organs could occur in the future. A comprehensive
follow up must continue for the lifetime of those exposed.
Oncogene (2009) 27, S9–S18; doi:10.1038/onc.2009.349
Keywords: Chernobyl; radiation; carcinogenesis; rearrangement; thyroid; latency
Introduction
The carcinogenic potential of ionizing radiation was
shown when skin cancers developed in early X-ray
workers, and Marie Curie herself developed leukemia
after working with radioactive isotopes. Many studies
have shown an increased cancer risk from exposure to
radiation—for example, the increased incidence of
second cancers in children treated by radiation for their
first malignancy. The detailed long-term studies of
atomic bomb survivors in Japan have made a major
contribution to the estimation of the radiation-doserelated risk of developing malignancy (Preston et al.,
2007).
In simple terms, ionizing radiation reacts directly or
indirectly with the genome of the radiated cell, leading
to one or more mutations. Those mutations leading to
activation of genes increasing cell growth, or inactivation of genes suppressing cell growth, can start the
progeny of the irradiated cell on the carcinogenic
pathway. This includes acquiring the additional mutations needed for a clinically detectable tumor. The
essential changes needed for malignancy include activation of an oncogene, loss of activity of both copies of a
tumor-suppressor gene and acquisition of the mutations
needed for invasion/metastasis. Five to seven has been
suggested as the minimum number of essential mutated
genes; recent work has shown that many more mutations are found in tumor cells, although the majority are
considered to be passengers rather than drivers. A study
of breast and colon tumors found an average of 90
mutated genes per tumor, 12 were thought to be
important for carcinogenesis (Sjöblom et al., 2006).
The large number may reflect a loss of DNA repair
activity—which also can result from radiation-induced
mutations.
This simple explanation for carcinogenesis is no
longer acceptable as the sole mechanism. Gene function
Radiation carcinogenesis: lessons from chernobyl
D Williams
S10
can be controlled by many mechanisms, including
epimutation, gene amplification, interaction with histones and other chromatin proteins. The recent demonstration that differentiated cells can be induced to revert
to stem cells by a virus-free technique not involving any
modification of their genome is also likely to be relevant
to carcinogenesis (Kaji et al., 2009).
Observations over the past two decades raise questions specifically relevant to radiation carcinogenesis.
Genomic instability describes the phenomenon that
irradiation may not lead to mutations in a cell, but
can lead to an increased risk of its daughter cells
developing chromosomal abnormalities or other mutations many cell generations later. The conclusion that
this effect is not because of a mutation induced at the
time of radiation is strengthened by the phenomenon
known as the bystander effect. Here, genomic instability
is observed in the progeny of an unirradiated cell
adjacent to an irradiated cell, suggesting humoral
transmission of instability. These observations do not
supplant the role of direct radiation-induced mutations
in DNA in radiation carcinogenesis, but they add other
mechanisms affecting gene function and other ways of
inducing mutations. They may be particularly important
in low-dose effects, and are relevant to the study of the
consequences of any major radiation accident.
The consequences of exposure to fallout after
Chernobyl cannot be predicted from atom bomb
studies. Some of the differences between the two events
are set out in Table 1. The type of radiation, dose rate
and the dose distribution in tissues, all differ. After
Chernobyl, trace amounts of radioactivity were detected
around the whole northern hemisphere; over 10 million
people lived in the most exposed areas of Belarus,
Ukraine and the Russian Federation, over two orders of
magnitude greater than the number of survivors of the
atomic bombs.
The accident
The accident at the No. 4 nuclear power plant at
Chernobyl in the far north of Ukraine occurred on 26
April 1986. It was the result of an ill-judged experiment
carried out while the plant was being shut down. The
reactor lacked the secondary containment found in
Table 1 Differences between the radiation exposure to the atomic
bombs in Japan and the radiation exposure from fallout after
Chernobyl
Type of radiation
Length of exposure
Tissues irradiated
Population
exposed
Oncogene
Atomic bomb
(Hiroshima and
Nagasaki)
Chernobyl fallout
Gamma rays and
neutrons
Instantaneous
Whole body
c105
Beta and gamma
Days to years
Variable with isotope
Over 107
modern nuclear power plants. Automatic safety devices
had been disabled; when the cooling system failed,
the pumps could not be restarted quickly enough. The
reactor started to go out of control, overheated,
the graphite core caught fire and a steam explosion
blew the lid off the reactor. The heat melted the fuel
rods, and the radioactive gasses and volatile isotopes
present in the fuel at the end of a cycle were released into
the atmosphere, while the molten nonvolatile radioactive isotopes penetrated through the reactor floor and
remained as a heap of slag in the basement. The fire and
release of radioactivity (over 1019 Bq) continued for over
a week. The wind initially carried the radioactive cloud
over southern Belarus, where the heaviest fallout
occurred. Parts of northern Ukraine and the adjacent
oblasts of what is now the Russian Federation were also
badly affected. During the days after the accident, the
radioactive cloud spread over large parts of Europe,
with the amount of fallout being dependent on
rainfall. Radioactivity from Chernobyl was detected as
far away as Japan. Heroic efforts were made to limit the
disaster, and many of the early workers (liquidators) at
the site received high doses of whole-body and isotopic
radiation. Of these workers, 32 died from acute
radiation syndrome within a few weeks, and the heavily
irradiated group continues to suffer from illnesses
similar to those seen after equivalent exposure to atomic
bomb radiation. However, the unprecedented events
after Chernobyl lie in the effects of radiation from
fallout on the very large number of the exposed general
population.
Incidence of thyroid carcinoma
The first indication of any increase in malignancy
among those exposed to fallout was observed in 1990,
only 4 years after the accident, when centers in Minsk
and Kiev noticed an increase in thyroid carcinoma in
children. There was initially skepticism from the West
with regard to the reports, but confirmation of the
clinical and pathological diagnosis led to two letters in
Nature, which triggered the start of extensive international collaboration (Baverstock et al., 1992; Kazakov
et al., 1992). There is now no doubt that isotopes of
iodine are the reason, about 1.7 1018 Bq of 131I was
released, second only to Xenon. Large amounts of 132Te,
rapidly decaying to 132I, were also released. The thyroid
has a unique ability to concentrate and bind radioactive
iodine, so that it receives a dose 500–1000 times higher
than the rest of the body.
The incidence of childhood thyroid carcinoma continued to increase, and in 1995 the incidence rate of
childhood thyroid carcinoma in Belarus reached 40 per
million (Demidchik et al., 1999). The incidence of
thyroid carcinoma varies in different countries; many
have an incidence of about one per million children per
year. The highest incidence was seen in the most exposed
oblast of Gomel in the south of Belarus, the lowest in
the least exposed oblast of Vitebsk in the north of the
Radiation carcinogenesis: lessons from chernobyl
D Williams
no. of cases
S11
12
80
70
60
50
40
30
20
10
0
10
8
6
4
<1 1
2
3
4
5 6 7 8 9 10 11 12 13 14
Age at exposure
2
Figure 1 Age at exposure and occurrence of thyroid cancer. All
childhood cases observed in Belarus between 1987 and 1997. The
slight increase in the oldest children could reflect the baseline
incidence (data from Cardis et al., 1999).
0
country. The report by WHO/IAEA in 2005, 20 years
after the accident (WHO, 2006), concluded that about
4000 thyroid cancer cases were caused by exposure to
fallout from Chernobyl, but it is difficult to define an
accurate figure, particularly as the incidence of thyroid
carcinoma has been rising steeply in unaffected countries. This increase is largely attributed to increasing use
of ultrasound and other techniques that detect small
thyroid lesions; increased ascertainment is enhanced
when both the general public and caring professions are
sensitized to the possibility of an increased risk. Another
problem arises from the paucity of direct measurements
of thyroid radioactivity immediately after the accident,
hence thyroid doses have to be reconstructed. However,
a strong dose relationship was found between radiation
exposure and the risk of developing thyroid cancer, with
an odds ratio at 1 Gy of 5.5, similar to the figure found
for external radiation (Cardis et al., 2005).
Age-related risk
Analysis of the early results from Belarus showed a very
strong link between young age at exposure and the risk
of developing thyroid carcinoma (Williams, 1996)
confirmed by later reports (Figure 1). After 1996, a
decline in incidence in children was accompanied by an
increase in adolescents; this in turn declined and an
increase was seen in young adults (Figure 2). Similar but
smaller increases were seen in the exposed areas of
Ukraine and the Russian Federation. An analysis of the
published data on thyroid cancers in Belarus occurring
in those under 19 years at the time of the accident
(Demidchik, 2005) shows that the peak age in those
operated in 1997 was in the 9–14 age group, whereas the
peak age for those operated in 2002 was in the 14–19 age
group. The slope and peak were very similar for both
years, suggesting that radiation-induced tumors show a
similar response to that seen in the first decade after the
accident. The results for all years analysed are shown
(Figure 3). The overlapping last points of the curves
suggest that they lie close to the baseline incidence,
which, as expected, rises with increasing age, whereas for
each year of operation, the numbers of radiation-related
88 89 90 91 92 93 94 95 96 97 98 99 0
Children
Adolescents
1
2
Young Adults
Figure 2 Change in incidence (per 105) of thyroid cancer over time
after the Chernobyl accident in 1986. The decline in the numbers of
children with thyroid cancer is accompanied by a rise in the
numbers of adolescents (data from YE Demidchik, in Cardis et al.,
2006)
90
80
70
60
50
40
30
20
10
0
-4,
-9,
-19,
-14,
-24,
Age at operation
-29,
-34,
1990
1991
1992
1993
1994
1995
1996
1997
1998
1999
2000
2001
2002
Figure 3 Numbers and ages of papillary thyroid carcinomas in
Belarus, in individuals aged less than 19 years at exposure, operated
from 1990 to 2002. The apparent peaks are because the ages are
shown in 5-year groups. The overlapping last points for each year
of operation come close to forming a smooth curve that is likely to
correspond approximately to the baseline incidence (data from YE
Demidchik, 2005)
tumors would be expected to continue falling as
increased age at operation equates to increased age at
exposure. Increased ascertainment has been estimated to
lead to a fourfold increase in baseline incidence in the
most exposed areas of Ukraine and Belarus (Heidenreich et al., 2004; Jacob et al., 2006). Although the
absolute risk of developing thyroid carcinoma after
exposure to fallout from Chernobyl may or may not
change with time, the relative risk will diminish with
increasing latency as the baseline incidence rises. Studies
of post-Chernobyl tumors should recognize that the
attributable fraction (the proportion of tumors due to
radiation) will be highest in those young at exposure
with short latency, and lowest in those older at exposure
with long latency.
The evidence so far presented suggests that the risk
for those exposed as adults is very low or absent. The
Oncogene
Radiation carcinogenesis: lessons from chernobyl
D Williams
S12
12
Incidence, x10-5
10
8
6
4
2
Gomel
Vitebsk
0
_15
_18
_35
_45
Age at operation
>45
Figure 4 Thyroid cancer incidence (1985–2006) in oblasts in
Belarus with the highest and lowest doses to the thyroid. The upper
line is from Gomel (average dose 0.44 Gy) and the lower from
Vitebsk (average dose 0.04 Gy). The incidence in the oldest age
group (all of whom were 25 years or more at the time of the
accident) is very similar, the major difference occurs in the groups
who were children at exposure (data from PI Bespalchuk et al.,
2009).
reported incidence of thyroid carcinoma in adults in
Belarus has risen sharply over the past two decades, but
before attributing this rise to radiation, the possibility of
increased ascertainment and of nonradiation-related
factors increasing incidence must be considered. Thyroid carcinoma incidence rates derived from tumor
numbers reported for Belarus (Bespalchuk et al., 2009)
show a similar incidence for those over 45 years for the
oblast with the highest exposure (Gomel, average
thyroid dose 0.44 Gy) and for the oblast with the lowest
exposure (Vitebsk, average thyroid dose 0.04 Gy). All
these cases would have been older than 25 years at the
time of the accident. The rates for Gomel were much
higher than those for Vitebsk for the three younger age
groups, which include all those under the age of 15 years
at the time of exposure (Figure 4). This provides further
evidence that the risk for those exposed as adults is very
low or absent.
Although many tissues show an age-related risk for
developing radiation-induced cancer, the difference
between children and adults is often a factor of
about 2. Two main factors must be considered to
explain the very much higher age-related sensitivity seen
with thyroid carcinoma: the dose to the child’s thyroid;
and biological factors influencing sensitivity. There
is no doubt that the radiation dose to the thyroid of a
child exposed to radioiodine in fallout is greater than
that of a similarly exposed adult. The main route
through which radioiodine in fallout is ingested is
through milk. Mammary epithelium concentrates iodine, both in cattle and in humans and milk forms
a much higher proportion of the diet in young
children than in adults. In addition, the thyroid in
young children tends to concentrate and bind iodine
more effectively.
Calculations of the risk per Gray for children have
suggested that radiation dose alone can explain the
Oncogene
difference in incidence between those exposed as young
and as older children (Jacob et al., 2006). This is an
unexpected finding. It is generally accepted that, for
adults over 40 years, the risk of developing thyroid
carcinoma after X-ray exposure is nonexistent or
negligible, and the observations discussed above suggest
that the same is true for internal radiation; there must
therefore be an earlier age-related change. After
exposure to external radiation, children under 4 years
of age showed a fivefold greater risk per Gy of
developing thyroid cancer than children aged 10–14
years (Ron et al., 1995) and a study of thyroid cancer
after radiotherapy for an earlier childhood cancer found
a 10-fold higher ERR/Gy for those treated with
radiation at age 0–1 years compared with ages 15–20
years (Ronckers et al., 2006). There are differences in the
dose rate and in the uniformity of dose distribution
across the thyroid between external and internal
radiation, which could possibly be relevant, but assessment of any age-related differences in dose-related risk
depends critically on the assessment of dose. This is
relatively simple for external radiation, but the methods
used after Chernobyl to reconstruct the thyroid dose for
children of differing ages were complex. It is possible
that these could underestimate the relative dose of
younger and older children.
There are biological reasons why it would be expected
that young children would have a greater sensitivity to
radiation-induced carcinogenesis. The thyroid grows
relatively rapidly during development; by the end of
adolescence, the growth rate is very low, and during
most of adult life, one in 1000 cells are in cycle at any
one time (Saad et al., 2006) and the thyroid mass is
constant. Point mutations are most likely to take place
during the S phase, and are therefore more likely to
occur in more rapidly dividing cells. Whether the first
radiation-induced mutation that leads eventually to the
development of thyroid carcinoma is a point mutation
or a rearrangement, additional mutations are needed for
progression to a clinically detectable tumor. The chance
that additional point mutations will occur is linked to
the number of cell divisions in the progeny of the
originally mutated cell. The adult thyroid is probably
not maintained by stem cells, and the growth capacity of
thyrocytes is limited (Wynford-Thomas et al., 1982).
The chance of acquiring additional mutations therefore
diminishes during childhood, and will be very low
during adult life. This mechanism can account for the
age-related change in the risk of developing thyroid
carcinoma after exposure to external radiation, and
would also apply to internal radiation. The estimates for
the thyroid dose in children after Chernobyl should be
reexamined.
Pathology and molecular biology
There are two main types of carcinomas derived from
the follicular cell, papillary and follicular. Both are
generally well differentiated. Undifferentiated carcinoma is highly aggressive and typically found in the
Radiation carcinogenesis: lessons from chernobyl
D Williams
S13
elderly. The earliest cases in the dramatic rise in the
incidence of thyroid carcinoma in children after
Chernobyl were all papillary carcinomas (PTC). In the
first few years, 98% of Belarussian and 94% of
Ukranian cases were PTCs (Williams, 1996). Over time,
small numbers of follicular carcinomas have occurred;
there is no proof that these are radiation induced, and
the proportion of papillary carcinomas in those less than
18 years of age at exposure in Belarus has remained
above 95% in most years. The incidence of thyroid
carcinoma in unexposed young children is so low that
virtually all early exposed cases in young children can be
assumed to be radiation induced.
PTCs show a variety of subtypes, and there have been
interesting changes in their frequency with time in the
exposed population. The earliest reports commented
that nearly all PTCs were of the solid subtype, and
speculated that this could be a marker of radiationinduced tumors. Later studies showed a decline in the
proportion of the solid subtype, and an increase in
the proportion of the classic subtype. Quantification of
the morphological changes showed that these changes
were significant; the less-mature solid tumors also
showed more direct invasion than the more mature
classic papillary carcinomas. Because the great majority
of tumors occurred in children who were very young at
exposure, the changes correlated both with increasing
age and increasing latency. Study of a group of tumors
in children who were older at exposure showed that
latency was the key factor (Williams et al, 2004).
Mutations in two major oncogenes have been shown
to be important in adult PTCs, point mutation in BRAF;
and rearrangement in RET. The two are almost entirely
mutually exclusive, a tumor is either BRAF or RET
positive. A minority show a TRK rearrangement; few
adult PTCs lack one of these changes. RET is normally
not expressed in the follicular cell; it is activated by
rearrangement, bringing the tyrosine kinase part of the
gene under the control of an active promoter. A variety
of genes can partner RET in the rearrangement, by far
the commonest are H4, resulting in RET–PTC1 and
ELE1, resulting in RET–PTC3.
Studies of early Chernobyl-related tumors by several
groups found that a very high proportion showed an
RET rearrangement, and almost all were RET–PTC3. It
was speculated that this rearrangement might be a
marker for radiation-induced tumors (Nikiforov et al.,
1997). Over time the proportion of tumors with an RET
rearrangement has declined, but in RET positive
tumors, the proportion with RET–PTC1 has increased
and the proportion with RET–PTC3 has decreased
(Rabes et al, 2000). TRK and RET–PTC2 rearrangements have been found in only a small proportion of
tumors; other RET–PTC rearrangements have been
described, usually in single cases.
In post-Chernobyl cases, BRAF point mutations are
uncommon (Lima et al., 2004), but the studies were
carried out in young patients. In unirradiated patients,
BRAF mutations are less frequent in childhood than in
adult PTCs. It remains possible that BRAF mutations
will become more common in ‘Chernobyl’ tumors as the
Possible course
6y latency
1992
14y
latency
2000
22y
latency
2008
30y
latency
2016
Baseline
Age at op
0
20
40
Type
Papillary
Papillary
Papillary
Subtype
Oncogene
Clinical
Latency
(Estimated,
Start-peak)
Solid
RET-PTC3
Aggressive
4-10
Solid/Class
RET-PTC1/3
Intermediate
Classical
RET-PTC1
Normal
7-15
? Papillary,
? Follicular or other
?
?RET,RAS,BRAF,PPAR
? depends on type
?
Figure 5 Hypothetical representation of the evolution of the
occurrence of thyroid carcinoma in those exposed to fallout after
Chernobyl. The baseline incidence is represented from age 0–40
years, and superimposed are the radiation-induced cases for
selected years, with correlation with morphology, oncogenes
involved and clinical behavior.
exposed population ages, but one study of thyroid
carcinomas in adults who had received external radiation found a low frequency of BRAF mutations (Collins
et al., 2006). It is also interesting that a small number of
cases with BRAF rearrangements have been found in
PTCs in children exposed to Chernobyl fallout (Ciampi
et al., 2005).
Several groups have investigated the relationship
between morphological subtypes of PTC and molecular
pathology. All have found that the solid type of PTC is
linked to RET–PTC3, and the classic type to RET–
PTC1 (Nikiforov et al., 1997; Thomas et al., 1999).
Two important conclusions can be drawn from these
studies. The first is the importance of latency; there have
been successive waves of tumors in those exposed to
high levels of fallout as children, each with different
molecular, morphological and clinical findings. It is
impossible to predict with certainty the further waves of
tumor that may occur. A diagrammatic representation
of the genotype—phenotype and latency–clinical behavior interaction is shown (Figure 5).
It is possible to construct a plausible scenario to
explain the pathobiology of the development of these
tumors. Although most spontaneous and chemical
carcinogen-induced mutations are point mutations,
radiation preferentially induces double-strand DNA
breaks, which can lead to deletions and rearrangements.
Although a variety of rearrangements may occur, some
will be lethal and only those that increase the chance of
further mutations are likely to contribute to carcinogenesis. RET–PTC3 has been shown in vitro to induce a
higher growth rate than RET–PTC1 (Soares et al.,
2003), and it may well be that this explains the shorter
latency and greater aggressiveness of RET–PTC3
tumors. The range of rearrangements induced by
radiation is also likely to be influenced by the interphase
arrangement of DNA; in the follicular cell nucleus, the
break points for RET–PTC1 have been shown to lie
Oncogene
Radiation carcinogenesis: lessons from chernobyl
D Williams
S14
very close together (Nikiforova et al., 2000). They
lie at the crossover point at which the DNA strand
forms a loop, so that a single ‘hit’ could break both
strands. If they are then joined wrongly, the inversion
that is seen in RET–PTC1 is formed. Attempts have
been made to identify a radiation-specific pattern of
gene expression using microarrays, but at present no
consensus exists.
It is interesting to speculate why follicular tumors,
benign or malignant, or PTCs with the typical BRAF
point mutation, have not yet been shown to increase in
incidence after exposure to Chernobyl fallout. The
commonest mutation found in follicular adenomas
is a point mutation in RAS; although PAX8-PPAR-g
rearrangements occur, in most cases, they are associated
with follicular carcinomas and are likely to represent a
late mutation during carcinogenesis. Follicular adenomas did increase in frequency after exposure to external
radiation, but with an extremely long latency (Shore
et al., 1993).
Role of rearrangements
The restriction of the increase in malignancy to the
thyroid is easily explained, but why the increase is
effectively restricted to one type of thyroid tumor and so
far to rearrangements involving one of two oncogenes
but not point mutations is more difficult, and raises the
general question of the role and specificity of rearrangements. Radiation is particularly effective at causing
double-strand breaks, and also induces point mutations
(Sankaranarayanan, 1991). In theory, the breaks could
occur anywhere in the genome, but several factors
influence the chance of the occurrence of a carcinogenic
rearrangement. In practice, there is differential susceptibility to double-strand breaks, and the chance of any
particular rearrangement occurring will be influenced by
the spatial arrangement of interphase DNA (Gandhi
et al., 2009). If the rearrangement does not contribute to
carcinogenesis, it is unlikely to be detected. Many
oncogenes such as RET and TRK code for a tyrosine
kinase, and the upstream gene in the rearrangement
must have a coiled-coil domain to result in nonliganddependent activation. Given these restrictions, it is even
more surprising that point mutations in BRAF or RAS
seem so far to have in the genesis of these thyroid
tumors.
The two classical short latency radiation-induced
tumors are thyroid carcinomas and leukemias. Both
very commonly show tumor type-specific rearrangements. In the thyroid, 15 different upstream partners
have been shown to be involved in rearrangements with
RET or TRK. It is noteworthy that five of the genes have
also been shown to fulfill a similar role in leukemias, in
combination not with RET or TRK but with other
oncogenes. Two of these five genes have been described
as candidate tumor-suppressor genes, and a third has
been implicated in ATM-mediated DNA repair. Two
more of the genes that are N-terminal partners of RET
in thyroid carcinomas have also been shown to have
properties suggesting that they may have a role in
carcinogenesis in addition to activating RET (Table 2).
Carcinogenesis requires a number of mutations, and a
single hit giving rise to a rearrangement that involves
both loss of function of one copy of a tumor-suppressor
gene and activation of an oncogene is more likely to
progress to a tumor and to be associated with a shorter
latency tumor than one that starts with a point
mutation. I would therefore suggest that the tumor
specificity of many rearrangements depends on the
combination of loss of function in one gene and gain
of function in another, with both selected because they
interact with the normal cellular pathways of the tissue
giving rise to the tumor. This analysis would suggest
that the explanation for the thyroid tumor findings after
Chernobyl starts with radiation-induced double-strand
breaks, with rearrangements influenced by an interphase
spatial arrangement of DNA and the selection of
carcinogenic rearrangements. The potent effect of
tyrosine kinases on follicular cell growth, the requirement for coiled-coil domains and particularly the tumorsuppressor or DNA repair gene function of the upstream partner determine the actual rearrangements
found. Potentially carcinogenic point mutations must
Table 2 Selected rearrangements found in papillary carcinoma of the thyroid
50 gene
30 gene thyroid
30 gene leukemia/
lymphoma
Potential contribution of 50 gene
Reference
H4 (D10S170)
RET
PDGFrB
PRKAR1A
RET
RARA
ATM pathway of DNA repair,
pro-apoptotic
Candidate tumor suppressor gene
PCM1
TFG
TPM3
TRIM24 (TIF1A)
TRIM27 (RFP)
RET
TRK
TRK
RET
RET
JAK2
ALK
ALK
Candidate TSG
Celetti et al. (2004);
Merolla et al. (2007)
Sandrini et al. (2002);
Griffin et al. (2004)
Armes et al. (2004)
Candidate TSG
Pro-apoptotic
Khetchoumian et al. (2007)
Dho and Kwon (2003)
Five of the 50 genes involved fulfill the same role in various leukemias or lymphomas. Three of these and two others display properties that when
inactivated could contribute to the role of the rearrangement in carcinogenesis. Each has other functions, and the role in carcinogenesis remains
unproven. Another (ELKS), like H4, is linked both to apoptotic and ATM pathways, but its role is uncertain (Wu et al., 2006).
Oncogene
Radiation carcinogenesis: lessons from chernobyl
D Williams
S15
have occurred, but these involve only one of the steps
needed to give rise to a clinically detectable tumor, and
are likely to have a longer latency than a rearrangement
involving two steps. Tumors with a point mutation as
the first event may well be found to be increased in the
future.
no specific genetic link has been shown. A number of
genes are known to be associated with the repair of
DNA double-strand breaks, among them BRACA
genes. In view of the link between thyroid and breast
cancer, the possibility of germline defects in BRACA, as
well as in other dsb repair genes such as ATM, needs
investigation.
Factors influencing the risk of thyroid carcinogenesis
Clinical outcome of thyroid carcinomas
The main risk factor is the radiation dose to the thyroid.
The Chernobyl accident is estimated to have released
about 1.7 1018 Bq of iodine 131, and although large
amounts of much shorter lived isotopes of iodine and
tellurium 132, decaying to iodine 132, were released,
these are of importance only for the population living
close to the reactor and exposed within a very short time
of the accident. Thyroid radioactivity was directly
measured in thousands of those exposed, but measurements were not made immediately after exposure;
release continued for about a week, and doses had to
be reconstructed from available data. Despite this, a
strong relationship was found between the dose received
in childhood and the subsequent risk of thyroid cancer,
with an odds ratio of 5.5–8.4 at 1 Gy in one study,
depending on the model used (Cardis et al., 2005). This
estimate is in the same range as that for the risk of
thyroid carcinogenesis after external radiation. The
importance of age at exposure and its interaction with
dose have been discussed.
Iodine status is also important. The areas around
Chernobyl are relatively iodine deficient, and in Belarus,
children from areas in the lower tertile of stable iodine
intake were found to have approximately three times the
risk of those in the upper tertile (Cardis et al., 2005).
Iodine deficiency affects the uptake of radioactive iodine
and therefore the dose to the gland, but most of the
effect depends on the increased size of the gland that
follows long-term iodine deficiency. There is evidence
that iodine deficiency may also influence the morphology and aggressiveness of childhood thyroid cancer
(Williams et al., 2008). Iodine deficiency is also a
possible factor in the unexpectedly short latent period
for the first increase in thyroid cancers after Chernobyl,
only 4 years after the accident.
Genetically determined susceptibility is potentially
important for the risk of developing thyroid cancer after
Chernobyl. Nonmedullary thyroid cancer has a relatively high familial element in nonradiated cases, but
although genes have been identified that are associated
with familial follicular and oxyphil tumors, and with the
special type of thyroid cancer associated with familial
adenomatous polyposis, no gene has yet been linked
with ‘ordinary’ papillary carcinomas. Polymorphisms
in DNA repair pathways have been shown to be
associated with thyroid cancer risk (Adjadj et al., 2009;
Bastos et al., 2009); in both studies polymorphisms in
XRCC3 were implicated. PTCs were found in exposed
siblings after Chernobyl more often than would have
been expected by chance (Cardis et al., 1999), but as yet,
The number of deaths in patients with thyroid
carcinoma exposed to high levels of fallout from
Chernobyl as children has been very low in the first 20
years; 15 deaths were reported in the WHO/IAEA
review (W). Deaths from causes other than thyroid
disease were not separately identified, nor were deaths
from medullary carcinoma. The aggressiveness of early
cases has been well documented, and in these cases,
direct extrathyroid invasion and lung metastases were
relatively frequent. An analysis of 740 cases of childhood thyroid cancer from Belarus found that 92% had
been exposed to Chernobyl fallout and 95% were PTCs
(Demidchik et al., 2006). Lymph node spread was found
in 69% of cases. A radioiodine scan showed that lung
metastasis was much more frequent than found by
conventional X-ray. Lung metastasis was strongly
associated with a young age at presentation. Five- and
10-year survival rates for the whole series were 99.5 and
98.8%, respectively. Of the eight patients who died,
three died of causes other than thyroid cancer, three
from medullary carcinoma and one from follicular
carcinoma. Only one patient died from papillary
carcinoma; this patient with widespread pulmonary
metastases was one of the early cases and was not
treated with radioiodine. These results underline the
generally good prognosis of appropriately treated
childhood papillary carcinoma, but of the 128 patients
with lung metastases, complete remission was achieved
in 29% cases, and those with incomplete remission are
still maintained on radioiodine therapy.
There is of course a risk of complications from
treatment. In the series just discussed, 6.2% suffered
permanent recurrent laryngeal nerve damage and 12.3%
suffered permanent hypoparathyroidism. These figures
are higher than those in adult series, but at least in part
reflect the difficulty of surgery in small children often
with local tumor spread. Second malignancies have been
reported after radioiodine therapy, including salivary
gland tumors. The salivary glands concentrate radioiodine, and are also close to the thyroid. One salivary
gland tumor occurred in a follow-up of 245 radioiodinetreated thyroid carcinomas from Belarus (Kumagai
et al., 2007). It is important to balance the risks and
benefits of high-dose radioiodine therapy for children
with thyroid carcinoma, and the risks include pulmonary fibrosis and secondary malignancies. Studies on the
very large cohort of childhood thyroid carcinomas after
Chernobyl should provide very valuable information to
help in these decisions.
Oncogene
Radiation carcinogenesis: lessons from chernobyl
D Williams
S16
With increasing latency, the patients are of course
older, and the tumors have in general become less
aggressive. It currently seems likely that the eventual
cause-specific death rate for the tumors that have occurred
in the first 20 years after the accident is less than 5%,
possibly much less. Hopefully this trend will continue for
cases that arise in the future, but it remains possible that
different mutations may lead to tumors with a different
morphology and different clinical behavior.
Other thyroid tumors
An increase in thyroid nodularity in more exposed areas
was an early finding, although interpretation was
complicated by the variable levels of iodine deficiency.
A Ukranian study has shown an increase in the
incidence of follicular adenomas in exposed children
and adolescents, with a linear dose–response relationship (Zablotska et al., 2008). The risk was lower than the
risk for carcinoma, but it must be remembered that the
mean latency for follicular adenomas after external
radiation was longer than that for carcinomas, hence the
risk ratios may change with time.
Nonthyroid diseases
Because of the early and dramatic increase in thyroid
carcinoma incidence, most of the attention on the health
consequences of the Chernobyl accident has been
focused on the thyroid. The nature of radiation
exposure suggests that nonthyroid consequences could
follow exposure to isotopes of iodine, or to other
isotopes including strontium, which is bone seeking, and
cesium, which is generally distributed throughout the
body. Cs 34 with a half-life of 30 years is still present at
above generally accepted safe levels in areas surrounding
the exclusion zone. The very high dose to the thyroid
from iodine 131 is dependent upon the ability of the
gland to concentrate, bind and store the isotope. A
variety of epithelia possess the iodide symporter,
allowing them to concentrate iodide, but the absence
of binding and storage means that the tissue dose is
significantly less than that of the thyroid. Mammary
epithelium, particularly when lactating, concentrates
iodine 131 and an increase in breast cancer in young
women in Gomel oblast has been reported (Pukkala
et al., 2006). It is not yet clear whether the increase is
related to lactation at the time of the Chernobyl
accident. The most susceptible time for radiationinduced breast cancer in atomic bomb studies was at
puberty, but it is uncertain whether pubertal breast
epithelium has a functional iodide transporter. The
possibility that breast cancer may also occur in the
future in post-Chernobyl thyroid cancer patients who
were treated at a young age with high doses of
radioiodine also needs further study.
A well-documented radiation-induced malignancy is
leukemia and radiation from isotopes of strontium
bound to bone could have contributed to the
Oncogene
marrow dose. Exposure to iodine 131 involves an
element of whole-body radiation, as does exposure to
other radioactive isotopes, especially cesium. Increases
in leukemia incidence in exposed populations have
been both claimed and denied; a recent assessment
concludes that, apart from cleanup workers, there is no
proof of a link to exposure (Howe, 2007). There have
been reports, sometimes anecdotal, of increases in a
variety of other tumors in those exposed to Chernobyl
fallout, including brain and kidney. In the absence of
thorough epidemiological studies, based on verified
diagnoses and taking into account the problems of
ascertainment, it is difficult to be certain that these are
Chernobyl-related increases. What they do show is the
continuing need for well-supported long-term studies
similar to those still being carried out after the atomic
bombs.
Germline effects
The possibility that radiation can lead to germ-cell
mutations that can be transmitted to subsequent
generations has been known for many years, but
has been associated with a relatively high-dose radiation. The finding of minisatellite instability in the
unexposed children of fathers exposed to Chernobyl
fallout (Dubrova et al., 1997) was unexpected and
concerning. There is as yet no evidence of any physical
disease in those carrying the instability. The dose to the
testis from fallout would be orders of magnitude lower
than the thyroid dose, unless there was an unexpected
concentration in the testis of one of the radioactive
isotopes released. Several authors have failed to find a
significant elevation of mini- or microsatellite mutations
in children of Chernobyl cleanup workers (Slebos et al.,
2004).
Conclusion
The accident at Chernobyl was unprecedented, millions of
individuals were exposed to radiation from isotopes in
fallout over a short period. This has allowed the
morphological and molecular changes in the resulting
tumors and their clinical behavior to be linked to latency in
a way that is not possible in ordinary practice. The
radiation exposure from Chernobyl fallout differed from
that from the atomic bombs, but long-term study
of the Japanese experience showed that many effects were
not apparent until decades after exposure. It is essential
that the consequences to human health of the Chernobyl
accident continue to be studied for the lifetime of those
exposed.
Conflict of interest
The author declares no conflict of interest.
Radiation carcinogenesis: lessons from chernobyl
D Williams
S17
References
Adjadj E, Schlumberger M, de Vathaire F. (2009). Germline
polymorphisms and susceptibility to differentiated thyroid cancer.
Lancet Oncol 10: 181–190.
Armes JE, Hammet F, deSilva M, Ciciulla J, Ramus SJ, Soo WK et al.
(2004). Candidate tumor-suppressor genes on chromosome arm
8p in early-onset and high grade breast cancers. Oncogene 23:
5697–5702.
Bastos HN, Antao MR, Silva SN, Azevedo AP, Manita I, Texeira V
et al. (2009). Association of polymorphisms of genes of the
homologous recombination DNA repair pathway and thyroid
cancer risk. Thyroid 19: 1067–1075.
Baverstock K, Egloff B, Pinchera A, Ruchti C, Williams D. (1992).
Thyroid cancer after Chernobyl. Nature 359: 21–22.
Bespalchuk PI, Demidchik YE, Demidchik EP, Saenko VA,
Yamashita S. (2009). Current trends in incidence and mortality
from thyroid cancer in Belarus. In: Nakashima M et al. (ed).
Radiation Health Risk Sciences. Springer: Tokyo.
Cardis E, Amoros E, Kesminiene A et al. (1999). Observed and
predicted thyroid cancer incidence following the Chernobyl
accident. In: Thomas G, Karaoglou A, Williams ED (eds).
Radiation and Thyroid Cancer. World Scientific: Singapore,
pp 395–405.
Cardis E, Kesminiene A, Ivanov V, Malakhova I, Shibata Y, Khrouch
V et al. (2005). Risk of thyroid carcinoma after exposure to 131I in
childhood. J Nat Canc Inst 97: 1–9.
Cardis E, Howe G, Ron E, Bebeshko V, Bogdanova T, Bouville A.
(2006) et al. Cancer consequences of the Chernobyl accident: 20
years after. J Radiol Protect 26: 127–140.
Celetti A, Cerrato A, Merolla F, Vitagliano D, Vecchio G, Grieco M.
(2004). H4(D10S170), a gene frequently rearranged with RET in
papillary thyroid carcinomas; functional characterization. Oncogene
23: 109–121.
Ciampi R, Knauf RA, Kerler R, Gandhi M, Zhu Z, Nikiforova MN
et al. (2005). Oncogenic AKAP9-BRAF fusion is a novel mechanism
of MAPK pathway activation in thyroid cancer. J Clin Invest 115:
94–101.
Collins BJ, Schneider AB, Prinz RA, Xu X. (2006). Low frequency
of BRAF mutations in adult patients with papillary thyroid
cancers following childhood radiation exposure. Thyroid 16:
61–66.
Demidchik EP, Mrochek A, Demidchik Yu, Vorontsova T, Cherstvoy
E, Keningsberg J et al. (1999). Thyroid cancer promoted by
radiation in young people of Belarus. In: Thomas G, Karaoglou A,
Williams ED (eds). Radiation and Thyroid Cancer. World Scientific:
Singapore, pp 51–60.
Demidchik Y. (2005). Thyroid cancer of Belarussians having been
exposed as children or adolescents as a result of the Chernobyl
accident. Report of the German Federal Office for Radiation
Protection, BMU-2005-668, appendix 8 1–20.
Demidchik YE, Demidchik EP, Reiners C, Biko J, Mine M, Saenko
VA et al. (2006). Comprehensive clinical assessment of 740 cases of
surgically treated thyroid cancer in children of Belarus. Ann Surg
243: 525–532.
Dho SH, Kwon KS. (2003). The Ret finger protein induces apoptosis
via its RING finger-B box-coiled-coil motif. J Biol Chem 278:
31902–31908.
Dubrova YE, Nesterov VN, Krouchinsky NG, Ostapenko VA,
Vergnaud G, Giraudeau F et al. (1997). Further evidence
for elevated human minisatellite mutation rate in Belarus
eight years after the Chernobyl accident. Mutation Res 381:
267–278.
Gandhi M, Evdokimova V, Nikiforov YE. (2009). Mechanisms of
chromosomal rearrangements in solid tumors: the model of
papillary thyroid carcinoma. Mol Cell Endocrinol, e-pub ahead of
print.
Griffin KJ, Kirschner LS, Matyakhina L, Stergiopoulos SG,
Robinson-White A, Lenherr SM et al. (2004). A transgenic mouse
bearing an antisense construct of regulatory subunit 1A of protein
kinase A develops endocrine and other tumors. J Med Genet 41:
923–931.
Heidenreich WF, Bogdanova TI, Biryukov AG, Tronko ND. (2004).
Time trends of thyroid cancer incidence in Ukraine after the
Chernobyl accident. J Radiol Protection 24: 283–293.
Howe GR. (2007). Leukemia following the Chernobyl accident. Health
Phys 93: 512–515.
Jacob P, Bogdanova TI, Buglova E, Chepurniy M, Demidchik Y,
Gavrilin Y et al. (2006). Thyroid cancer among Ukrainians and
Belarusians who were children or adolescents at the time of the
Chernobyl accident. J Radiol Protection 26: 51–67.
Kaji K, Norrby K, Paca A, Mileikovsky M, Mohseni P, Woltjen K
et al. (2009). Virus free induction of pluripotency and subsequent
excision of programming factors. Nature 458: 715–716.
Kazakov VS, Demidchik EP, Astakhova LN. (1992). Thyroid cancer
after Chernobyl. Nature 359: 21.
Khetchoumian K, Teletin M, Tisserand J, Mark M, Herquel B, Ignat
M et al. (2007). Loss of Trim24 (Tif1alpha) gene function confers
oncogenic activity to retinoic acid receptor alpha. Nat Genet 39:
1500–1506.
Kumagai A, Reiners C, Drozd V, Yamashita S. (2007). Childhood
thyroid cancers and radioactive iodine therapy. Endocr J 54:
839–847.
Lima J, Trovisco V, Soares P, Máximo V, Magalhães J, Salvatore G
et al. (2004). BRAF mutations are not a major event in postChernobyl childhood thyroid carcinomas. J Clin Endocrinol Metab
89: 4267–4271.
Merolla F, Pentimalli F, Pacelli R, Vecchio G, Fusco A, Grieco M
et al. (2007). Involvement of H4 (D10S170) protein in ATMdependent response to DNA damage. Oncogene 26: 6167–6175.
Nikiforov YE, Rowland JM, Bove KE, Monforte-Munoz H, Fagin
JA. (1997). Distinct pattern of ret oncogene rearrangements in
morphological variants of radiation-induced and sporadic thyroid
papillary carcinomas in children. Cancer Res 157: 1690–1694.
Nikiforova MN, Stringer JR, Blough R, Medvedovic M, Fagin JA,
Nikiforov YE. (2000). Proximity of chromosomal loci that
participate in radiation-induced rearrangements in human cells.
Science 290: 138–141.
Preston DL, Ron E, Tokuoka S, Funamoto S, Nishi N, Soda M et al.
(2007). Solid cancer incidence in atomic bomb survivors:1958–1998.
Radiat Res 168: 1–64.
Pukkala E, Kesmininene A, Poliakov S, Ryzhov A, Drozdovitch V,
Kovgan L et al. (2006). Breast cancer in Belarus and Ukraine after
the Chernobyl accident. Int J Cancer 119: 651–658.
Rabes HM, Demidchik EP, Siderow JD, Lengfelder E, Beimfohr C,
Hoelzel D et al. (2000). Pattern of radiation induced RET and
NTRK1 rearrangements in 191 post Chernobyl papillary carcinomas: biologic, phenotypic and clinical implications. Clin Cancer Res
6: 1093–1103.
Ron E, Lubin JH, Shore RE, Mabuchi K, Modan B, Pottern LM et al.
(1995). Thyroid cancer after exposure to external radiation, a
pooled analysis of 7 studies. Radiat Res 141: 259–277.
Ronckers CM, Sigurdson AJ, Stovall M, Smith SA, Mertens AC, Liu
Y et al. (2006). Thyroid cancer in childhood cancer survivors.
Radiat Res 166: 618–628.
Saad AG, Kumar S, Ron E, Lubin JH, Stanek J, Bove KE et al.
(2006). Proliferative activity of human thyroid cells in various age
groups and its correlation with the risk of thyroid cancer after
radiation exposure. J Clin Endocrinol Metab 91: 2672–2677.
Sandrini F, Matyakhina L, Sarlis NJ, Kirschner LS, Farmakidis C,
Gimm O et al. (2002). Regulatory subunit type 1-alpha of protein
kinase A (PRKARIA): a tumor-suppressor gene for sporadic
thyroid cancer. Genes Chromosomes Cancer 35: 182–192.
Sankaranarayanan K. (1991). Ionizing radiation and genetic risks.
Mutat Res 258: 75–97.
Shore RE, Hildreth N, Dvoretsky P, Pasternack B, Andresen E.
(1993). Benign thyroid adenomas among persons X-irradiated in
infancy for enlarged thymus glands. Radiat Res 134: 217–223.
Oncogene
Radiation carcinogenesis: lessons from chernobyl
D Williams
S18
Sjöblom T, Jones S, Wood LD, Parsons DW, Lin J, Barber TD et al.
(2006). The consensus coding sequences of human breast and
colorectal cancers. Science 314: 268–274.
Soares P, Trovisco V, Rocha AS, Lima J, Castro P, Preto A et al.
(2003). BRAF mutations and RET/PTC rearrangements are
alternative events in the etiopathogenesis of PTC. Oncogene 22:
4578–4580.
Slebos RJ, Little RE, Umbach DM, Antipkin Y, Zadaorozhnaja TD,
Mendel NA et al. (2004). Mini- and microsatellite mutations in
children from Chernobyl accident cleanup workers. Mutation Res
559: 143–151.
Thomas GA, Bunnell H, Cook HA, Williams ED, Nerovnya A,
Cherstvoy ED et al. (1999). High prevalence of RET-PTC
rearrangements in Ukranian and Belarussian post-Chernobyl
thyroid papillary carcinomas. J Clin Endocr Metab 84: 4232–4238.
WHO. (2006). Health Effects of the Chernobyl Accident and Special
Health Care Programmes; Report of the UN Chernobyl Forum
Expert Group ‘‘Health’’ (EHG). World Health Organisation:
Geneva.
Williams ED. (1996). Effects on the thyroid in populations exposed to
radiation as a result of the Chernobyl accident. In: One Decade after
Oncogene
Chernobyl. International Atomic Energy Authority: Vienna, pp
207–230.
Williams ED, Abrosimov A, Bogdanova T, Demidchik EP, Ito M,
LiVolsi V et al. (2004). Thyroid carcinoma after Chernobyl,
latent period, morphology and aggressiveness. Br J Cancer 90:
2219–2224.
Williams ED, Abrosimov A, Bogdanova T, Demidchik EP, Ito M,
LiVolsi V et al. (2008). Morphological characteristics of Chernobylrelated childhood papillary thyroid carcinomas are independent
of radiation exposure but vary with iodine intake. Thyroid 18:
847–852.
Wu ZH, Shi Y, Tibbets RS, Miyamoto S. (2006). Molecular linkage
between the kinase ATM and NF-KappaB signalling in response to
genotoxic stimuli. Science 311: 1141–1146.
Wynford-Thomas D, Stringer BMJ, Williams ED. (1982). Desensitisation of rat thyroid to the action of TSH. Acta Endocrinol 101:
562–569.
Zablotska LB, Bogdanova TI, Ron E, Epstein OV, Robbins J,
Likhtarev IA et al. (2008). A cohort study of thyroid cancer and
other thyroid diseases after the Chernobyl accident. Amer J
Epidemiol 167: 305–312.