Cancer Risks Following Low Dose Radiation Exposures Lessons

  • Slides: 23
Download presentation
Cancer Risks Following Low Dose Radiation Exposures: Lessons from Epi Studies The Accidents at

Cancer Risks Following Low Dose Radiation Exposures: Lessons from Epi Studies The Accidents at Fukushima Dai-Ichi Exploring the impacts of Radiation on the Ocean November 13, 2012 Dale L. Preston Hirosoft International Eureka, CA

Topics • Describing long-term radiation effects on disease risks • Follow-up studies of radiation

Topics • Describing long-term radiation effects on disease risks • Follow-up studies of radiation health effects – Atomic bomb survivors (Life Span Study) – Techa river residents • Results – Leukemia – Solid cancers • Conclusions 2

Radiation Effects on Disease Risks • Affected cases indistinguishable from other cases • Magnitude

Radiation Effects on Disease Risks • Affected cases indistinguishable from other cases • Magnitude of effect depends on dose – Dose rate may be important – Effects (if any) of low to moderate doses appear to be “small” • Effects can appear and persist long after exposure • Effects can depend on sex, age, and other factors • Detection and characterization requires: – – Large exposed populations (especially when doses are low) High quality long-term follow-up Good dose estimates Careful analyses using sophisticated statistical methods, 3

Disease Risks and Rates • Risk – Probability of disease occurrence during a specified

Disease Risks and Rates • Risk – Probability of disease occurrence during a specified period (e. g. lifetime after exposure, within 10 years of exposure, …) • Rate – Ratio of the number of new cases in a time period to the total time at risk in the time period (person-years) – Risks can be computed from rates – Characterization of rates is central to modern description of radiation effects. 4

Disease rates and radiation effects • Baseline rate B 0 – Describe population rates

Disease rates and radiation effects • Baseline rate B 0 – Describe population rates in the absence of radiation exposure – Depends on age, sex (s), and other factors • Describing rates in an exposed population – Relative risk model: B 0 [ 1 + ERR(d)] • The excess relative risk (ERR) describes the magnitude of the radiation effect relative to the baseline rate. – Rate difference model: B 0 + EAR(d) • The excess absolute rate (EAR) describes the difference between the rate in the exposed an unexposed population. 5

Describing Radiation Effects on Rates Issues • Shape of dose response – Linear (LNT)

Describing Radiation Effects on Rates Issues • Shape of dose response – Linear (LNT) – Non-linear (quadratic, J-shaped, high dose attenuation, …) – Threshold • Dose rate effects • Effect modification – Time since exposure, age at exposure, attained age, sex, ethnicity, … • Interactions with other risk factors – Additive – Multiplicative – Other 6

Atomic Bomb Survivor Studies • Initial studies (1947 -53) focused on genetic effects in

Atomic Bomb Survivor Studies • Initial studies (1947 -53) focused on genetic effects in children of survivors – No evidence of heritable genetic effects • Anecdotal reports of radiation-associated leukemia from late 1940’s – Leukemia registry established in early 1950’s • Life Span Study (LSS) cohort established for long-tern follow-up of survivors – All-cause mortality and leukemia incidence follow-up since 1950 – Solid cancer incidence follow-up since 1958 • Clinical follow-up of a subset of the LSS since 1958 7

Life Span Study • 120, 321 people with 93, 741 exposed – 58% women

Life Span Study • 120, 321 people with 93, 741 exposed – 58% women – 40% exposed as children • Individual organ dose estimates for 93% of exposed – – Mean weighted colon dose ~170 m. Gy within 3 km Maximum 4000+ m. Gy 25% of exposed 5 -100 m. Gy 15% 100+ m. Gy • Virtually complete follow-up – Since 1950 for mortality and leukemia incidence • 10. 929 solid cancers; 371 leukemias – Since 1958 for solid cancer incidence • 17, 448 cases (3, 994 unexposed) 8

LSS Solid Cancer Incidence 1958 -98 Dose Response • Linear ERR per 100 m.

LSS Solid Cancer Incidence 1958 -98 Dose Response • Linear ERR per 100 m. Gy 0 – 2 Gy 0. 05 • No evidence of non-linearity (LQ model on 0 – 2 Gy) P > 0. 5 • 11% (850/7, 851) of cases among those with 5 m. Gy or more associated with radiation exposure 48% above 1 Gy - 307/645 Using RERF public dataset lssinc 07. csv (www. rerf. or. jp) Proximal zero dose baseline (adjusted for distal and NIC) 9

LSS Solid Cancer Incidence Dose Response 0 – 0. 5 Gy • Linear ERR

LSS Solid Cancer Incidence Dose Response 0 – 0. 5 Gy • Linear ERR per 100 m. Gy 0 – 2 Gy 0. 05 0 – 100 m. Gy 0. 05 • LSS often described as a high dose study, but has as more information on risk at relatively low doses (<100 m. Gy) than many low dose studies Test for trend 0 – 100 m. Gy P = 0. 08 Using RERF public dataset lssinc 07. csv (www. rerf. or. jp) Proximal zero dose baseline (adjusted for distal and NIC) 10

LSS Solid Cancer Incidence Temporal Patterns ERR EAR • Significant effect modification by attained

LSS Solid Cancer Incidence Temporal Patterns ERR EAR • Significant effect modification by attained age and age at exposure and sex (ERR only) 11

LSS Leukemia Risks 1950 -2001 • Analyses based on 312 cases – CLL (12

LSS Leukemia Risks 1950 -2001 • Analyses based on 312 cases – CLL (12 cases) and adult T-cell leukemia (47 cases) were excluded • Significant non-linear dose response ERR at 1 Gy 2. 3 (age 60 age at exposure 25) ERR at 100 m. Gy 0. 09 • No significant sex-difference in the ERR 12

LSS Leukemia Temporal Patterns • ERR decreases with both attained age with no sex

LSS Leukemia Temporal Patterns • ERR decreases with both attained age with no sex difference • EAR decreases with attained age and time since exposure with lower risks for women than men 13

Mayak Plutonium Production Association • Secret facility located in a closed territory in the

Mayak Plutonium Production Association • Secret facility located in a closed territory in the Southern Urals • Began operation in 1948 • Produced Pu used in first Mayak Soviet nuclear weapon 14

Mayak Production Association • Complex occupational exposures – External gamma – Inhaled plutonium for

Mayak Production Association • Complex occupational exposures – External gamma – Inhaled plutonium for radiochemical and plutonium production worker • ~40% monitored for Pu exposure – Highest doses in 1948 -53 period • Multiple Environmental exposures 15

Mayak-related Environmental Releases • Discharges into Techa River 1950 -56 100 PBq Cs 137,

Mayak-related Environmental Releases • Discharges into Techa River 1950 -56 100 PBq Cs 137, Sr 90, Sr 89, …. (98% 1950 -51) • Khyshtym accident 1957 70 PBq Rare earths, Sr 90 Y 90 • Karachai resusupension 1967 35 PBq Cs 137, Sr 90 • Gaseous aerosols 1948 -55 40 PBq I 131 16

Techa River Cohort • 29, 730 residents of 41 riverside villages born before 1950

Techa River Cohort • 29, 730 residents of 41 riverside villages born before 1950 – 58% women – 39% exposed as children • Individualized organ dose estimates – 35 m. Gy mean stomach dose (8. 5 m. Gy per year 1950 -53) – 414 m. Gy mean marrow dose (72 m. Gy per year 1950 -53) (90% from Sr) • Follow-up – Since 1950 for morality; 1953 for leukemia incidence; 1956 for solid cancer incidence – ~20% lost to follow-up due to migration – 2, 303 solid cancer deaths; 99 leukemia cases (27 CLL) 17

Techa River Solid Cancer Mortality • ERR at 100 m. Gy 0. 06 (95%

Techa River Solid Cancer Mortality • ERR at 100 m. Gy 0. 06 (95% CI 0. 004 to 0. 13) • No significant non-linearity, but power to detect non-linearity is low • Comparable to LSS estimate No significant sex difference in ERR • • • Weak suggestion that ERR increases with increasing age 2% (50/2303) of solid cancer deaths associated with exposure Significant dose-response following low dose rate exposures ERR broadly similar to that seen in the LSS Little power to characterize effect modification or to look at causespecific solid cancer mortality 18

Techa River Leukemia Risks • Significant dose response with ERR per 100 m. Gy

Techa River Leukemia Risks • Significant dose response with ERR per 100 m. Gy of 0. 23 • No significant non-linearity • No significant attained age, sex, or ethnicity effects • Weak suggestion that doses received 2 to 10 years before diagnosis have higher ERR (0. 5) than doses received more than 10 years earlier (0. 17) – Similar pattern seen in Mayak workers – Pattern consistent with LSS temporal variation • 47% (34 of 72) of cases associated with radiation exposure 19

Conclusions • Epidemiological studies of radiation exposed populations are challenging and expensive, but have

Conclusions • Epidemiological studies of radiation exposed populations are challenging and expensive, but have provide important information on the risks at low doses and low dose rates • The exposures of Techa River residents are similar in nature to those received by people in areas contaminated by the Fukushima accident • Studies of the Techa River cohort find clear evidence of low dose rate radiation effects on cancer risks • Techa River risks are comparable to LSS risk and provide no evidence of a reduction in effect at low doses 20

Fukushima Challenges • A well-designed long-term epidemiological study should be developed – Power to

Fukushima Challenges • A well-designed long-term epidemiological study should be developed – Power to detect effects may be limited but failure to find effects would be a reassuring message • Studies should include an assessment of the long term psychological effects – These have not been investigated in a systematic manner in other major radiation exposed cohorts – Could help in developing better ways to work with public in the even of future radiation accidents 21

Sources LSS Solid cancer incidence Preston et al Radiat Res 2007 168: 1 -64

Sources LSS Solid cancer incidence Preston et al Radiat Res 2007 168: 1 -64 LSS Leukemia incidence Hsu et al Radiat Res in press Preston et al Radiat Res 1995 137(Suppl 2): 68 -97 Techa Solid cancer Schonfeld et al Radiat Res in press Krestinina et al Radiat Res 2005 164: 502 -611 Krestinina et al Int J Epidemiol 2007 36: 1038 -1046 Techa Leukemia Krestinina et al Leukemia (tent. ) Krestinina et al Radiat Environ Biophys 2010 49: 195 -2001 22

Acknowledgments I am grateful for long term support of my efforts in the Atomic

Acknowledgments I am grateful for long term support of my efforts in the Atomic bomb survivor and Southern Urals studies provided by the National Cancer Institute, the Department of Energy, and the Radiation Effects Research Foundation I have benefitted greatly from my collaborations with: Elaine Ron (NCI), Don Pierce (RERF), Ludmila Krestinina and Marina Degteva (URCRM – Techa River), Sara Schonfeld (NCI, IARC) and many more 23