Hormesis and Risk Estimate Comparisons and Applications in Radiation Protection

1. Hormesis andRisk Estimate Comparisonsand Applications inRadiation Protection

2. Radiation Hormesis Early literature used “hormoligosis”, derived from the Greek hormo (meaning “to excite”) and oligo (meaning “low doses”) The shortened “hormesis” is preferred The concept of “dose is everything” described as early as Hippocrates, recognizing that small stresses stimulate while excess stresses inhibit

3. Radiation Hormesis Exposure to UV rays of sun activate production of Vitamin D. A few minutes of exposure results in daily allowance of Vitamin D. UV light used to cure some dermatologic conditions Prolonged exposure results in sunburn and increased incidence of skin cancer Some would argue that even a minute of exposure may cause cancer. Risk is evident at cellular level but how and when, if at all, does the dose become detrimental?

4. Radiation Hormesis

5. Radiation Hormesis The Zero Thesis: “All radiation is harmful” The Hormesis Thesis “small and large doses produce opposite results “Radiation hormesis in mammals is readily observed when reasonable constraints are used to examine the data” T.D. Luckey, 1991 “The dose is everything” Paracelsus, 1540

6. Radiation Hormesis Some things to consider: Fight or Flight response of biologic systems: stressors stimulate At “hormetic” doses cellular repair to DNA damage is very efficient Hormesis can not be predicted by extrapolation from higher, harmful radiation doses

7. Hormesis and Dose/Response

8. Zero Equivalence Point

9. Viscoelasticity of Liver DNA

10. Additional “Proof” “Test of LNT Theory of Radiation Carcinogenesis for Inhaled Radon Products” (Cohen, 1995) Used data for 1601 US counties to test LNT theory Corrections for widely recognized problems with ecologic studies e.g. smoking prevalence, geography, altitude, socioeconomic Still showed strong tendency for lung cancer rates to decrease with increasing radon exposure Found no explanation other than failure of LNT

11. More Proof Incredible Hulk – mutated by atomic blast perceived hormesis Godzilla – also mutated by Atomic blast. Appearance in early 1950’s perhaps comment on socio-psychologic issues of US dropping atomic bombs in Japan. Monster not altogether unfriendly although occaisionally reaks havoc with populations. Spider man – bit by radioactive spiderIncredible Hulk – mutated by atomic blast perceived hormesis Godzilla – also mutated by Atomic blast. Appearance in early 1950’s perhaps comment on socio-psychologic issues of US dropping atomic bombs in Japan. Monster not altogether unfriendly although occaisionally reaks havoc with populations. Spider man – bit by radioactive spider

12. On an Cellular Level Low level radiation activates four damage control mechanisms damage prevention damage repair damage removal by apoptosis damage removal by a stimulated immune response Excerpts from: SOME NON-SCIENTIFIC INFLUENCES ON RADIATION PROTECTION STANDARDS AND PRACTICE by Laursiton S. Taylor   [The original article appeared in Health Physics 1980; 32, pp 851-874. The excerpts were chosen by John Cameron with verbal approval from the author.] Excerpts from: SOME NON-SCIENTIFIC INFLUENCES ON RADIATION PROTECTION STANDARDS AND PRACTICE by Laursiton S. Taylor   [The original article appeared in Health Physics 1980; 32, pp 851-874. The excerpts were chosen by John Cameron with verbal approval from the author.]

13. Cellular Mechanisms Low doses of radiation result in increased: Membrane permeability Protoplasmic streaming Respiration Enzyme synthesis Metabolism Bioelectric potential Mitosis High doses result in the opposite effects

14. DNA Repair Mechanisms The quantity of cellular enzymes involved in DNA repair produced as a result of irradiation is greater than that needed to repair the damage from the radiation itself Therefore, there is a surplus of repair enzymes which can make the cell more radio-resistant for future exposures In addition “hit” cells may communicate with neighboring cells to surplus these repair enzymes

15. Mechanisms Liver enzymes show increased activity, which in turn plays a role in increased growth rate, of “lightly” exposed animals Immune system can be stimulated by low doses such that white blood cell production is increased --> more efficient infection fighting --> longer life

16. Luckey’s Literature Review Acceptable for determination of hormesis if: Growth & survival studies must begin w/ young animal Adequate controls must be available Data from unchallenged, protected colonies is questioned Reports must give original data & straightforward statistics Sick control animals cannot be excluded Extrapolation from high doses is not acceptable Extrapolation from animals to humans is reasonable Cells in culture cannot mimic complex physiology Given these criteria, over 90% of the literature was excluded from Luckey’s analysis

17. Luckey’s Findings Growth and Development Reproduction Immunity Cancer

18. Growth and Development “Ample data” to support growth hormesis “Experiments which showed no radiation hormesis in growth are usually found to have started with young adults” Dose rates on the order of a few cGy/day resulted in increased weight gain in young rabbits, mice, guinea pigs, and fish.

19. Reproduction Many articles describe increased rates of mitosis, colony growth, litter size Fertility of trout shown to increase with low dose irradiations (25 - 100 cGy acute) No excess mutations in 82 generations of male parent (mice) receiving 2 Gy at 26 days old

20. Immunity Resistance to radiation shown in mice “pre-exposed” to doses ~10 times less than the “test” dose Small incisions made in rabbits & dogs following exposure showed faster healing than controls - high doses retarded healing Stimulated proliferation of bone marrow stem cells in mice exposed to 1-20 cGy x-rays

21. Cancer Luckey says BIER V’s straight-line stochastic model is incorrect, that “stochastic” as a word is obsolete and refers to conjecture Comparison of background radiation and cancer incidence shows “hormesis was evident” Says there are many exposures with no visible effects Page 104 in LuckeyPage 104 in Luckey

22. Cancer - ABS Acute doses were not separated properly from chronic doses States that a suitable control population does not exist and that medical radiations of survivors is not taken into account Points out # of cancers in the “0” exposed cohort of the ABS are less than expected Page 150 of Luckey with observed/expectedplotsPage 150 of Luckey with observed/expectedplots

23. Final Word from ICRP

24. BEIR V Somatic Risks Relative projection model Linear-quadratic dose-response model for leukemia and linear model for all others DREF of 2 calculated using leukemia data, but not utilized anywhere Specific calculations for leukemia, thyroid, respiratory, digestive, and “other” cancers

25. Other Risk Estimates Other committees all utilize same datasets, but make different assumptions and come to different conclusions BEAR & BEIR UNSCEAR NCRP, ICRP EPRI EPA, NRC NRPB

26. BEIR III Somatic Risks Prior to BEIR V, thought of as the “last word” Somatic risks - projection and dose-response models allowed to vary so as to “bound” the estimates. Dose-response model accounts for both gamma and neutron components separately Linear response assumed for neutrons, while LQ best represents gamma component Terms allowed to vary as in Monte Carlo which bounds the risk estimates. Gamma – linear quadratic model used while linear response assumed for neutronsTerms allowed to vary as in Monte Carlo which bounds the risk estimates. Gamma – linear quadratic model used while linear response assumed for neutrons

27. BEIR III Model BEIR III considered 5 exposure scenarios: Acute - 0.1 Gy Chronic - 0.01 Gy/yr lifetime Chronic - 0.01 Gy/yr from 20-65 yrs Chronic - 0.01 Gy/yr from 35-65 yrs Chronic - 0.01 Gy/yr from 50-65 yrs Dose-response model results put into a lifetable to consider gender, age, latency, expression, and projection

28. BEIR III Model Each of these models used to bound the risk estimate: Variations in equations used. Estimates also bounded by model used. Variations in equations used. Estimates also bounded by model used.

29. Somatic Risk Summary Close in relative and absolute projection of acute vs chronic risk.Close in relative and absolute projection of acute vs chronic risk.

30. BEIR III and V BEIR V risk estimates higher because: No use of DREF in BEIR V Linear model for non-leukemias New dosimetry 10 yrs of additional followup Changes in model structure Changes in model structure – LQ for Leukemias only. Only five types of cancer with no specifc organ calculations also only three exposure scenarios.Changes in model structure – LQ for Leukemias only. Only five types of cancer with no specifc organ calculations also only three exposure scenarios.

31. BEIR I Somatic Risks Prior to BEIR III there was BEIR I Only about 20 yrs of ABS followup Linear model utilized Risks based on those cancers that were appearing quickly (expression unknown)

32. Genetic Risks BEIR V uses linear model, with doubling dose for equilibrium estimates and direct method for 1st generation estimates BEIR III uses a DD range of 0.50 - 2.5 Gy (determined from a single mouse experiment showing that DD = 1.14 Gy) BEIR I uses DD range of 0.2 - 2.0 Gy

33. Genetic Risk Summary DD Spontaneous rates can differ greatly but genetic risks on same order of magnitude.DD Spontaneous rates can differ greatly but genetic risks on same order of magnitude.

34. Genetic Risk Summary (Cont’d)

35. Genetic Risk Summary (Totals)

36. Recall Dose (D) --> energy per unit mass Equivalent Dose (H) --> H = D • QF Effective Dose (E) --> E = S (HT • wT) Quality factor describes radiation type and its potential biological effect Tissue weighting factor accounts for relative organ radiosensitivity

37. Tissue Weighting Factors ICRP 26 first developed wT from BEIR III risk estimates: 7 organs used fraction of total risk as weighting factor7 organs used fraction of total risk as weighting factor

38. Tissue Weighting Factors With BEIR V, ICRP 60 attempts to refine wT They include factors for: Years of life lost Non-fatal contributions (morbidity) ICRP 60 weighting factors indicative of “detriment” rather than fatality alone Detriment includes non-fatal cancers, fatal cancers and years of life lost/cancer typeDetriment includes non-fatal cancers, fatal cancers and years of life lost/cancer type

39. Tissue Weighting Factors Number of organs increases from 6 to 13 (plus genetic), including:

40. Years of Life Lost Handled by normalizing estimates for each cancer type to the average (15 yrs) For example, bladder cancer deaths result in an average of 9.8 yrs of life lost, so its normalizing factor is 0.65 (9.8/15) Bone marrow deaths result in 30.9 yrs of life lost and a factor of 2.06 Bone marrow deaths 3 times more “detrimental” to society 15 yrs is average life lost for cancer15 yrs is average life lost for cancer

41. Morbidity Handled by calculating a “relative non-fatal contribution” from the factor, k, which is the fraction of fatalities resulting from a given cancer type For example, leukemia (bone marrow) has a “lethality fraction”, k, of 0.99 meaning that only about 1% of those with radiogenic leukemia will survive

42. Relative Morbidity

43. Morbidity The factor (k-2) used as a “normalizing” factor to account for non-fatal detriment Values range from 1 (highly lethal cancers) to 2 (highly curable cancers)

44. ICRP 60 Detriment Factors

45. Aggregate Detriment The total detriment is calculated as the product of the fatal cancer risk, the relative length of life lost, and the relative non-fatal contribution

46. Relative Contribution to Detriment

47. Organ Detriment Put in bins of 1, 5, 12, or 20 for convenience and to illustrate uncertaintyPut in bins of 1, 5, 12, or 20 for convenience and to illustrate uncertainty

48. Uncertainty Because of uncertainties in the method, ICRP 60 establishes 4 weighting factors -- 1, 5, 12, & 20% with organ sensitivities put into these bins Thus, the wT from ICRP 60 gives an indication of organ risk (detriment), but should not be used to estimate relative risks

49. ICRP 60 Risk Factors Same magnitude as Beir VSame magnitude as Beir V