Radiological Risk from Low Dose and Low Dose-Rate Exposures: An - PowerPoint PPT Presentation

Radiological Risk from Low Dose and Low Dose-Rate Exposures: An Epidemiologic Perspective MELODI Conference Munich, November 2015 Roy Shore, Linda Walsh, Werner Rühm, Tamara Azizova hrshore@gmail.com (Abridged version)

Disclaimer This represents preliminary concepts and data, and not necessarily the position of the ICRP Task Group-91 on Dose and Dose-Rate Effectiveness Factors. 2

DDREF: Dose and Dose-Rate Effectiveness Factor (DDREF) Questions for Low-LET radiations Is there a need for a DDREF? If so, …  How large is the DDREF?  Should the DDREF be separated into a:  LDEF (low-dose effectiveness factor) – upward curvature in the dose-response function, rather than linear, for a single brief exposure  DREF (dose rate effectiveness factor) – less effect per unit dose for low dose rates or numerous small exposures, or …  Are the values of the LDEF and DREF the same? 3

Recent Views regarding DDREF  From radiobiological studies DDREF estimates often have ranged from 2 to >5.  ICRP (2007) and UNSCEAR (2006) indicated compatibility with a DDREF of 2  BEIR VII (USA, 2006) derived an estimate of DDREF of 1.5 (with a potential range of 1.1-2.3)  SSK (Germany, 2014) – report concluded that no DDREF is necessary (i.e., DDREF of 1). 4

Low-Dose Effectiveness Factor (LDEF): A-bomb Life Span Study 5

Life Span Study (LSS) Solid Cancer Mortality Risk: Linear & Linear-Quadratic Dose Response • The quadratic term was statistically significant over the dose range 0-2 Gy, with estimated curvature ( β / α ) = 0.8 • Significant dose response on 0-200 mGy Excess Relative Risk (ERR) Weighted Absorbed Colon Dose (Gy) (Ozasa, Radiat Res , 2012;177:229-) 6

LSS dose response: Solid-cancer incidence  “ No significant evidence” of non-linearity ( p =0.09) in the dose response  Estimated curvature ( β / α ) = 0.3 (90% CI 0.01, 0.9)  Significant dose response on 0-150 mGy (LSS Incidence, 1958-1998) 150 % Excess Relative Risk ERR/Gy= 47% (90%CI: 40-54%) Dose-threshold: 40 mGy (90% CI: <0, 85 mGy) 100 50 Fitted linear dose response at age 70 following exposure at age 30 Smoothed non-parametric dose response 0 0 1 2 3 Weighted Absorbed Colon Dose (Gy) (Preston D et al: Radiat Res 168:1-64, 2007)

LSS solid cancer incidence (1958-98) (data from Preston, Radiat Res, 2007) Linear ERR@ 1Gy: 1.5 1.5 Linear 0.47 (95% CI 0.38, 0.56) Bayesian Semi-parametric: 0.45 (95% CI 0.36, 0.56) Excess relative risk 1.0 1.0 Excess relative risk 0.5 0.5 0.12 0.10 0.08 Excess relative risk 0.06 0.0 0.04 0.0 0.02 0.00 -0.02 0.00 0.05 0.10 0.15 0.20 DS02 weighted colon dose (Gy) 0.0 0.5 1.0 1.5 2.0 2.5 3.0 0.0 0.5 1.0 1.5 2.0 2.5 3.0 DS02 weighted absorbed colon dose (Gy) (Furukawa K, Risk Analy , In press, 2015) DS02 weighted colon dose (Gy) 12

LSS solid cancer incidence (1958-98) 0-200 mGy Linear 1.5 0.12 0.12 0.10 0.10 Bayesian Semi-parametric Excess relative risk 0.08 0.08 1.0 Excess relative risk 0.06 0.06 0.04 0.04 0.5 0.02 0.02 0.00 0.00 0.0 -0.02 -0.02 0.00 0.05 0.10 0.15 0.20 0.0 0.0 0.5 0.5 1.0 1.0 1.5 1.5 2.0 2.0 2.5 2.5 3.0 3.0 0 50 100 150 200 DS02 weighted colon dose (Gy) (Furukawa K, Risk Analy , In press, 2015) DS02 weighted absorbed colon dose (mGy) 13

DREF: Some Issues in Addressing Risk after Low-Dose or Low Dose-Rate (LDLDR) Exposures 10

Low-Dose Studies: Intrinsic Implications  As the Radiation Signal-to- Background (“ signal:noise ”) ratio of cancers decreases at lower doses, radiation effects become increasingly uncertain  Therefore, both statistical power and statistical precision (per unit dose) decrease greatly at low doses, i.e., uncertainty increases appreciably .  Bias from unmeasured confounding variables (e.g., other disease risk factors) has the potential to be relatively greater in a low-dose study  Because the magnitude of bias (either to exaggerate or mask the true degree of association) may approach or exceed the magnitude of the dose effect. 11

Sample Size Needed to Study Various Doses with Adequate Statistical Power, Lifetime Risk (Brenner et al, PNAS 100:13762-, 2003) 12

Issues to Consider in Analyzing and Interpreting LDLDR Studies  Potential biases due to dose-dependent differences, e.g.  Lifestyle factors – e.g., smoking, alcohol consumption  Healthy Worker Effects  Unaccounted for radiation exposures – e.g., medical  Dose-related differential adequacy of diagnosis and completeness of cancer ascertainment  Sources of uncertainty, e.g.,  Dosimetric uncertainties  Quality of the cancer morbidity or mortality data  Uncertainties in modeling, risk transport to other populations, etc 13

Examining the Dose-Rate Effectiveness Factor (DREF) in Low-Dose or Low Dose-Rate (LDLDR) Studies 14

Approach to Examine Low Dose and/or Low Dose Rate (LDLDR) Studies  Conduct a meta-analysis. Why?  Provide “weight of evidence”.  Much larger numbers of cancers and person-years needed at low doses to detect risks and achieve precise risk estimates.  Inclusion criteria: Studies must have a dose-response risk estimate Gy -1 and have low doses and/or highly fractionated or protracted exposures  Compile comprehensive list of studies with dose-response analyses of LDLDR data – try to avoid redundancy among studies, use latest data, and minimize study selection biases (e.g., publication bias, “cherry picking”)  Meta-analysis to evaluate ratios of risk coefficients in the LDLDR studies to the matching LSS data – (cf. Jacob et al, Occup Environ Med , 2009) 15

Rationale to Examine DREF for Specific Cancer Types Radiation effects for various types of cancer may be modified differently by:  Tumor biology: relevant genes & genetic pathways; epigenetic, tissue and metabolic cofactors  Other environmental risk factors – e.g. impact of smoking, alcohol intake, infections.  Baseline rates which vary across populations and over time. A knowledge gap exists regarding organ specific risks, especially following acute or protracted exposures under a few hundred mSv. 16

Summary of Preliminary Findings regarding Low Doses  Low Dose Effectiveness Factor – preliminary evidence of upward curvature for solid cancer in the A-bomb LSS  All solid cancer DREF: Most (9/11) of the LDLDR studies with >250 cancers had positive risk coefficients  Meta-analyses are underway to compare LDLDR risk with LSS risk  LDLDR summaries of specific tumour sites:  Breast cancer DREF – substantial evidence for LDLDR risk  Lung cancer DREF – weak evidence of LDLDR risk. But problem of possible confounding by smoking  Stomach cancer DREF – moderate evidence of LDLDR risk  Colon cancer DREF – small numbers of studies and cancers. Little evidence of risk 17

Epidemiologic Gaps in Knowledge  Uncertainties in the risk of leukemia and solid cancers at low doses (<100, <50, <20 mSv) and low dose rates (but data are becoming stronger)  Are low-dose risk and DDREF factors similar for various tumor sites?  Are low-dose radiation risks modified by disease risk factors: lifestyle, infectious, reproductive, etc?  How large an impact do biological and genetic susceptibility factors have on the radiation induction of cancer or cardiovascular disease, especially at low doses? 18

Epidemiologic Data and DDREF Assessment  Epidemiologic data regarding DDREF are important because they directly model human populations that are highly heterogeneous with respect to innate susceptibility and exposure co-factors.  Ultimate judgements regarding DDREF, however, will need to integrate information about associated biological mechanisms, experimental studies of dose and dose-rate factors in controlled animal experiments, and the epidemiologic observations. 19