Carcinogenic Effects of Low Doses of Ionizing Radiations

1
Carcinogenic Effects of Low Doses of Ionizing
Radiations

• R.Julian Preston
• Environmental Carcinogenesis Division
• U.S Environmental Protection Agency
• Research Triangle Park, NC
• AAAS
  Meeting, 2004

2
Types of Dose-response Curves
3
THE CHALLENGE
How to estimate human tumor responses at
low doses.
4
THE REQUIREMENT
The best fit tumor dose response curve has to be
best for the total range of tumor types and for
different radiation scenarios.
5
THE APPROACHES

• Extrapolate from human data at
• higher doses.
• Use animal tumor data
• Use surrogates for cancer responses-
• cellular and molecular indicators

6
What is a Low Dose?

• The background radiation levels are about
• 1mSv (0.001Sv) not including radon.
• Cancer risk levels can be computed for
• this level of radiation that for the present
• discussion is considered to be a low dose.

7
Examples of Low-Dose Exposures

• Round-trip flight, NY to London 0.1 mSv
• Background radiation (no radon) 1.0 mSv
• Background radiation with radon 3.0 mSv
• Single screening mammogram 3.0 mSv
• Pediatric abdominal CT scan 25 mSv
• International Space Station (per yr)170 mSv
• Radiation Worker yearly limit 20 mSv

8
Cancer Risk for Ionizing Radiation

• Based on human tumor data
• A-bomb survivors
• Radiation Workers
• Medical exposures
• Assumes linear-no threshold (LNT) response
• for extrapolation to low, environmental exposures

9
Radiation Cancer Risk Estimates

• Fatal cancer 4 per Sv
• Nonfatal cancer 0.8 per Sv
• Background frequency 25

10
CANCER EPIDEMIOLOGY

• (1) A-bomb survivors
• Doses less than 0.5 Sv.
• 7000 tumors 50,000 survivors.
• Reliable risk estimates to 0.05-0.1 Sv.
• Not overestimated by linear
• extrapolation from estimates computed
• over higher dose ranges.

11
CANCER EPIDEMIOLOGY

• (2) Medically exposed groups.
• Fractionated and protracted
• exposures can increase total
• cancer risk.
• Clear evidence of risk not
• available for low exposures
• limited precision, large errors.

12
CANCER EPIDEMIOLOGY
Conclusion

• Finite risk at low dose for solid tumors.
• Plausible for leukemias.
• Low-dose linearity is the best estimate
• of dose responsiveness.

13
ANIMAL CARCINOGENICITY

• Few generalizations can be drawn.
• Variability of responses for different species,
  
• strains, tumor types, radiation types.
• All types of dose-response curve shape
  observed.
• Radiation-induced life-shortening generally
• linear with dose.
• Low dose linearity cannot be ruled out for the
• great majority of studies similar
  generalization
• cannot be made for other curve shapes.

14
ANIMAL CARCINOGENICITY
Conclusion
Overall LNT is the best estimate of dose-response
for radiation- induced tumors in animal models.
15
CELLULAR AND MOLECULAR BIOLOGY STUDIES

• Cancer is a multi-step process.
• Cancer results from mutations or chromosomal
• alterations in oncogenes and tumor suppressor
• genes.
• Dose-response information for radiation-
• induced mutations and chromosome
• alterations is informative to tumor dose
• response.

16
GENE MUTATIONS

• Radiation-induced mutations increase
• as a linear, quadratic or
  linear-quadratic
• function of dose.

17
CHROMOSOMAL ALTERATIONS

• 99.9 of all tumors contain chromosomal
• alterations
• Radiation-induced chromosomal alterations
• can predict tumor response at low doses, in a
  
• qualitative sense.
• Two DNA lesions required.
• Produced by one or two tracks (low LET).
• Produced by one track (high LET).
• Significant increases over background at
• 100 mGy.
• For low LET, dose-response curves are
• linear or linear quadratic.

18
Cellular Responses to Ionizing Radiation
Potential Modifiers of Low Dose Cancer Risk

• Bystander effects
• Genomic instability
• Adaptive responses
• Note Current cancer risk estimates for radiation
  are based on human tumor frequencies and so
  incorporate any influence of these responses

19
Conclusions

• Bystander responses and induced genomic
  instability could lead to enhanced tumor
  responses at low doses.
• Adaptive responses could lead to a reduction in
  tumor responses at low doses
• Underlying mechanisms need to be better
  understood.

20
SUMMARY STATEMENT
Based upon information for radiation-induced human
tumors, animal tumors and surrogates for
tumors, LNT is frequently the best fit to the low
dose data and generally cannot be ruled out.
LNT is the best solution
To-day. However, more needs to be done and is
being done. The arguments presented are for
To-day.
21
Bystander Effects
Mutation
Mutation
Mutation
22
Bystander Effects

• Mediated by cell signalling events either cell
• cell communication or possibly through the
• tissue culture medium.
• Very little evidence for in vivo responses.
• Difficult to demonstrate for chemicals- except
• by single cell injection or using medium
• transfer experiments

23
Bystander Effects

• Impact on risk assessment ?
• Increases target cell population for a given
• exposure how can dose be defined in
• biological terms?

24
Genomic Instability

• Tumor cells generally show extensive genomic
  instability (structural and numerical).
• Early stage event selection for specific
  phenotypes drives the cancer process.
• Following radiation exposures (in vivo and in
  vitro) chromosomal alterations and mutations can
  be observed many cell divisions later delayed
  response. Described as genomic instability.
• Need to distinguish limited response in
  experimental systems with extensive response in
  tumor cells. Are these the same event?

25
Genomic Instability

• Induced instability results in higher mutagenic
  response than predicted on the basis of initial
  cellular response.
• Important for dose-response considerations- what
  is the response per unit dose?
• Could lead to supralinear response at low doses
  for tumor outcomes.
• Very limited evidence for genomic instability
  following chemical exposures.

26
Adaptive Response

• Small priming dose can reduce response to larger
  challenge dose.
• Demonstrated for chemical and radiation exposures
  for a range of endpoints, e.g. mutation,
  chromosome alterations.
• Reduces responses in most cases by a factor of
  about 2.

27
ADAPTIVE RESPONSE

• Does an adaptive response lead to a
• threshold for tumor induction?
• Adaption only reduces the response
• does not eliminate it.
• No evidence for adaptive response at
• very low dose rates.

28
ADAPTIVE RESPONSE
Conclusion
No version of an adaptive response necessitates
departure from LNT (based on cellular studies) a
change in slope can be predicted.
29
Interactions of Responses

• Adaptive responses and genomic instability
• have been reported to be induced in bystander
• cells, further complicating dose-response
• assessment.