BACKGROUND:We previously reported a significantly increased risk of thyroid cancer due to radioactive iodine-131 after the 1986 Chernobyl nuclear accident in a cohort of children and adolescents from Belarus. Radiation-related cancer risks were five times higher among those with diffuse goiter; however, the relationship between radiation dose and diffuse goiter is not well understood. METHODS:We used logistic regression to analyze data from 10,278 study participants [mean thyroid dose = 0.63 gray (Gy)] who were screened 10 to 15 years after exposure according to a standardized protocol, which specified thyroid size assessment by both ultrasound and palpation. RESULTS:Diffuse goiter was identified in 1,811 subjects and was significantly associated with 131I dose (P < 0.001). The linear model provided the best fit to the data at doses <0.5 Gy [odds ratio (OR) at 1 Gy = 2.45; 95% confidence interval, 1.70-3.41] and linear-exponential model at doses <1.5 Gy (OR at 1 Gy = 1.84). Age at exposure and indicators of iodine deficiency both before the accident (place of residence) and during screening (urinary iodine levels) significantly modified the dose-response (all P values < 0.01). CONCLUSIONS:The first systematic evaluation of radiation risks of diffuse goiter after environmental exposure found a strong, statistically significant association with thyroid dose. IMPACT:Our study shows that radiation exposure is associated with a significantly increased risk of developing diffuse goiter. Future epidemiologic studies of iodine-deficient irradiated populations should assess thyroid size.

BACKGROUND:Cataract is a well-established radiogenic condition and recent evidence suggests an increased risk even at radiation doses of <100 mGy. Following the 2011 Fukushima Daiichi Nuclear Power Plant accident, ∼20 000 workers participated in emergency operations. This study aimed to investigate the association between cumulative low-dose occupational radiation exposure and the risk of cataract among workers enrolled in the Epidemiological Study of Health Effects in Fukushima Nuclear Emergency Workers. METHODS:Data from 5773 nuclear emergency workers participating in the cohort were analysed. The cumulative lifetime occupational radiation dose was estimated by combining pre-2011 individual dose records from a nationwide nuclear radiation worker dose registry with radiation doses received during the 2011 emergency-work period. Self-reported, physician-diagnosed cataract was identified through baseline and first follow-up questionnaires. A Cox proportional hazards model was applied to estimate hazard ratios (HRs) with 95% confidence intervals (CIs) for incident cataract, with adjustments for major covariates. RESULTS:More than 90% of the participants had cumulative occupational radiation exposure of <100 mSv, including doses received both prior to the accident and during the emergency period (March-November 2011). During 52 921 person-years of follow-up from 2012 to 2024 (mean, 9.2 years), 310 participants reported new-onset cataract. The risk of cataract increased with lifetime occupational exposure up to November 2011, with an HR of 1.03 per 10-mSv increase (95% CI, 1.02, 1.05). This association persisted after excluding participants with doses of ≥100 mSv. For exposure during the emergency period (March-November 2011), the HR was 1.06 per 10-mSv increase (95% CI, 1.03, 1.09). When the cumulative lifetime occupational radiation dose was recalculated by using 2-year and 5-year lag periods before diagnosis or censoring, the HRs per 10-mSv increase were 1.03 (95% CI, 1.01, 1.04) and 1.01 (95% CI, 0.99, 1.04), respectively. CONCLUSION:This study suggests an excess risk of cataract at doses of <100 mSv and adds to the evidence base suggesting an association between low-dose occupational radiation exposure and cataract risk.

BACKGROUND:The large and growing burden of keratinocyte carcinoma (KC), including basal cell (BCC) and squamous cell carcinoma (SCC) suggests a need to identify high-risk individuals. We examined the association between KC risk and history of blistering sunburns in childhood and adulthood and investigated whether sun sensitivity modifies these associations. METHODS:Data were obtained from the large, nationwide U.S. Radiologic Technologists cohort, with baseline sunburn history and sun sensitivity traits collected via self-administered questionnaires. Participants were followed until diagnosis of first primary cancer or completion of a follow-up questionnaire (2003-2014). Hazard ratios (HR) and 95% confidence intervals (95%CIs) were estimated using Cox proportional hazards models. RESULTS:Among 40,204 participants, those reporting ever blistering sunburn had higher risk of BCC (HR=1.37;95%CI:1.26,1.49) and SCC (HR=1.41;95%CI:1.19,1.66). The HR of BCC per blistering sunburn in childhood and adulthood were 1.06 (95%CI:1.05,1.07) and 1.05 (95%CI:1.04,1.06) respectively. The HRs for SCC were 1.08 (95%CI:1.06,1.11) and 1.07(95%CI:1.05,1.09) per blistering sunburn in childhood and adulthood, respectively. Most sun sensitivity factors and other measures of UVR exposure did not significantly modify these associations. CONCLUSIONS:Number of blistering sunburns was associated with increased KC risk independently of an individual's pigmentary features and ambient UVR. The associations were similar for childhood and adulthood sunburns. IMPACT:History of blistering sunburn is an important risk factor for KC, even among groups that might otherwise be overlooked due to their overall lower risk of KC, such as individuals who have darker pigmentary features or those residing in locations with less ambient UVR.

BACKGROUND:While breast cancer risk from high-dose ionising radiation is known, uncertainties remain about risks at lower doses and risk-modifying factors. We conducted a systematic review and meta-analysis of publications on radiation-associated risk of breast cancer in women. METHODS:We included studies published in 2005-2022 that assessed breast cancer incidence or mortality in women exposed to ionising radiation. Risk of bias was evaluated. Random-effects meta-analyses estimated excess relative risks per gray (ERR/Gy). RESULTS:Of the 3522 articles screened, 106 met the inclusion criteria; 40 studies provided 44 ERR/Gy estimates. Overall, radiation exposure was associated with increased breast cancer risk (ERR/Gy = 0.56, 95% CI: 0.29-0.83). Between-study heterogeneity (I2 = 95%) was substantially reduced in subgroup analyses, reaching 5% in low-dose-rate studies. Higher summary ERRs were observed for high dose-rate exposures, moderate (1-5 Gy) doses, childhood exposures, and attained age over 55. Lower but significantly increased risks were estimated for other subgroups and exposure scenarios. CONCLUSIONS:Radiation exposure was associated with a significantly increased risk of breast cancer among women, particularly following high dose-rates, moderate doses, childhood exposures, and older attained age. PROSPERO registration: CRD42021260610.

The U.S. Radiologic Technologists (USRT) study investigates cancer and other serious disease risks associated with low-dose occupational radiation exposure. The previous dose system (URDS13) for the full cohort was based on badge dose records through 1997, three self-reported questionnaires administered between 1983 and 2005, and historical estimates from the literature. In this article, we describe an extended (23 additional calendar years, 1998-2020), updated and enhanced dosimetry system for the USRT cohort (URDS25). We incorporated 1,156,584 newly acquired annual badge dose readings (1977-2020) obtained by integrating annual summary data (1977-2011) and monthly badge reading data (2004-2020), bringing the total to 1,416,420 annual badge dose readings (1960-2020) for 81,885 technologists. To enhance the individualization of dose estimates, we also utilized detailed work history and protection practice data from the fourth survey, administered in 2012-2013, along with supplementary work history modules for technologists who performed nuclear medicine and assisted with fluoroscopically guided interventional procedures. Based on all badge readings and work history data collected to date, we re-evaluated the URDS13 badge dose estimates prior to 1997 and reconstructed the estimates from 1998 to 2020, resulting in a total of 3.27 million estimated annual badge doses for 110,374 technologists for the years 1916-2020. Each annual badge dose was reconstructed as a probability distribution using Monte Carlo simulation, generating 1,000 realizations to account for uncertainty in the true dose. Compared to the previous version (URDS13, 1916-1997), this update (URDS25, 1916-2020) resulted in a slight increase in the mean cumulative dose estimates for the entire cohort, from 76 mSv (median: 47 mSv; range: 0.19-3,000 mSv) to 79 mSv (median: 48 mSv; range: 0.12-3,000 mSv), and a mean absolute change of the individual cumulative dose estimates of 26% per technologist. Organ absorbed doses will also be updated based on these revised badge dose estimates and detailed, self-reported work history information and, along with updated follow-up data, will be used in future dose-response analyses to more precisely investigate radiation-associated cancer and non-cancer disease risks.

BACKGROUND:Few epidemiological studies have distinguished the effects of solar ultraviolet (UV) radiation (UVR) wavelength, including UVB and UVA, on the risk of basal cell carcinoma (BCC) and squamous cell carcinoma (SCC) of the skin. OBJECTIVES:We aimed to evaluate the association between ambient UVB and UVA exposure and the incidence of BCC and SCC. METHODS:Using data from the nationwide US Radiologic Technologists cohort, satellite-based noontime ambient UVR was obtained based on residential history (< 13, 13-19, 20-39, 40-64, and ≥ 65 years). Incidence rate ratios (IRRs) and 95% confidence intervals (CIs) were estimated for associations between UVB, UVA quintile (Q), and first primary BCC and SCC in mutually adjusted and stratified models, additionally adjusted for sun sensitivity characteristics. RESULTS:There were 62 595 non-Hispanic White participants aged 22-90 years at baseline (1983-1998). During a median 25.5 years of follow-up, 6339 incident BCCs and 1253 incident SCCs were reported. Annual and summer UVB and UVA were each positively associated with both BCC and SCC risk before mutual adjustment. After adjusting for UVA, summer (but not annual) UVB Q5 vs. Q1 was associated with BCC (IRR 1.41, 95% CI 1.13-1.77) and SCC (IRR 1.69, 95% CI 1.01-2.84). After adjusting for UVB, annual (but not summer) UVA Q5 vs. Q1 was associated with BCC (IRR 1.42, 95% CI 1.13-1.77) and SCC (IRR 1.84, 95% CI 1.10-3.06). CONCLUSIONS:Both summer UVB and annual UVA were independently associated with higher risks of both BCC and SCC. Our results suggest that cumulative lifetime UVA exposure may be an underappreciated contributor to BCC and SCC risk. If confirmed, these findings may indicate that public health interventions are required to promote avoidance of excessive UVB and UVA exposure among susceptible populations.

Objective: To establish thyroid ultrasound volume norms appropriate for studies of diffuse goiter in a cohort of children and adolescents from an iodine-deficient population exposed to 131I by the Chernobyl fallout. Methods: A cohort of 11,970 Belarusians aged ≤18 years at the time of the 1986 Chernobyl accident with individual thyroid radiation dose estimates was screened 10–18 years later. From these, a low-dose subset of 2,392 with no thyroid diseases was selected to construct age- and sex-specific normative values for thyroid ultrasound volume, compared to Belarusian Ministry of Health (MOH) norms and existing WHO and European standards. Results: Cohort-specific values were generally lower than MOH norms and WHO standards for 11–17-year-olds. For those aged ≥18 years, internal norms were 30% higher in males and 15–30% lower in females than MOH norms, and exceeded European values for both sexes. Thyroid volume norms were about 40% higher in males and 30% higher in females as a function of BSA compared to European values. Thyroid volume continued to increase in both sexes, and by age 30–34 years, cohort-specific norms were 6% higher in males and 26% higher in females than European values. Urinary iodine concentration did not significantly explain variance in thyroid volume beyond sex, age, and BSA. Conclusions: In this iodine-deficient cohort of young Belarusians exposed to 131I by Chernobyl fallout, thyroid ultrasound volumes differed substantially from MOH norms and established WHO standards, prompting a revision of diffuse goiter definition using cohort-specific normative values.

We investigate methods that improve the estimation of indoor gamma ray dose rates at locations where measurements had not been made. These new predictions use a greater range of modelling techniques and larger variety of explanatory variables than our previous examinations of this subject. Specifically, we now employ three types of machine learning models in addition to the geostatistical, nearest neighbour and other earlier models. A large number of parameters, mostly describing the characteristics of dwellings in the area in question, have been added to the set of explanatory variables. The use of machine learning methods results in significantly improved predictions over earlier models. The machine learning models are noisy and there is some instability in the relative importance of particular explanatory variables although there are general and consistent tendencies supporting the importance of certain classes of variable. However, the range of predicted indoor gamma ray dose rates is much smaller than that of the measurements. It is probable that epidemiological studies using such predictions will have lower statistical power than those based on direct measurements.

PURPOSE:Low-dose radiation risks are generally extrapolated from groups exposed at high levels of dose. Measurement error substantially alters dose-response shape and hence extrapolated risk. Much attention has been paid to methods of dealing with shared errors, common in many datasets, in particular using Bayesian model averaging (BMA) methods. MATERIALS AND METHODS:We test two types of BMA model using simulated data. The first we term the quasi-two-dimensional Monte Carlo with BMA (quasi-2DMC + BMA) method, similar to the BMA method proposed by Hoeting et al. but distinct from the 2DMC + BMA method of Kwon et al. The second we term the marginal-quasi-2DMC + BMA method, which uses a more complicated marginal calculation, and may be closer to the 2DMC + BMA method of Kwon et al. RESULTS:Assuming a true linear model of dose response, the coverage probabilities for the linear coefficient are 90-95%, for the quasi-2DMC + BMA method, but lower than this, 52-60%, for the marginal-quasi-2DMC + BMA method. Assuming a true linear-quadratic model the coverage probabilities of both linear/quadratic excess relative risk (ERR) coefficients for quasi-2DMC + BMA are too low, and when shared Berkson error is large (50%) the probabilities do not exceed 5%. By comparison, the coverage probabilities for both linear and quadratic coefficients for the marginal-quasi-2DMC + BMA method are generally too high, about 100%. For the linear model quasi-2DMC + BMA yields good estimates of the ERR coefficient, but for marginal-quasi-2DMC + BMA these are upwardly biased. For the linear-quadratic model both quasi-2DMC + BMA and marginal-quasi-2DMC + BMA methods yield substantially biased estimates. CONCLUSIONS:The performance of both quasi-2DMC + BMA and marginal-quasi-2DMC + BMA methods is bad, with bias and poor coverage.

PURPOSE:High-linear energy transfer (LET) radiation is generally thought to be more biologically effective in various tissues than low-LET radiation, but whether this also applies to the circulatory system remains unclear. We therefore reviewed biological studies about the effects of high-LET radiation on the circulatory system. CONCLUSIONS:We identified 76 relevant papers (24 in vitro, 2 ex vivo, 51 in vivo, one overlapping). In vitro studies used human, bovine, porcine or chick vascular endothelial cells or cardiomyocytes, while ex vivo studies used porcine hearts. In vivo studies used mice, rats, rabbits, dogs or pigs. The types of high-LET radiation used were neutrons, α particles, heavy ions and negative pions. Most studies used a single dose, although some investigated fractionation effects. Twenty-one studies estimated the relative biological effectiveness (RBE) that ranged from 0.1 to 130, depending on radiation quality and endpoint. A meta-analysis of 6 in vitro and 8 in vivo studies (selected based on the feasibility of estimating the RBE and its uncertainty) suggested an RBE of 6.69 (95% confidence intervals (CI): 2.51, 10.88) for in vitro studies and 1.14 (95% CI: 0.91, 1.37) for in vivo studies. The meta-analysis of these 14 studies yielded an RBE of 2.88 (95% CI: 1.52, 4.25). This suggests that high-LET radiation is only slightly more effective than low-LET radiation, although substantial inter-study heterogeneity complicates interpretation. Therapeutic effects have also been reported in disease models. Further research is needed to better understand the effects on the cardiovascular system and to improve radiation protection.

Biological effects of ionizing radiation vary with radiation quality, which is often expressed as the amount of energy deposited per unit length, i.e., linear energy transfer (LET). For acute irradiation, high-LET radiation generally produces greater biological effects than low-LET radiation, but little knowledge exists as to how dose protraction modifies effects. In this regard, inverse dose protraction effects (IDPEs) are phenomena in which dose protraction enhances effects, contrasting with sparing dose protraction effects in which dose protraction reduces effects. Here, we review the current knowledge on IDPEs of high-LET radiation. To the best of our knowledge, since 1967, 80 biology or epidemiology papers have reported IDPEs following external or internal high-LET irradiation with neutrons, deuterons, α-particles, light ions, or heavy ions. IDPEs of high-LET radiation have been described for biochemical changes in cell-free macromolecules, neoplastic transformation, cell death, DNA damage responses and gene expression changes in mammalian cell cultures of human or rodent origin, gene mutations, cytogenetic changes, cancer, non-cancer effects (e.g., testicular effects, cataracts, cardiovascular diseases) and life shortening in non-human mammals (rodents and dogs), and induction of lung cancer and bone tumors in humans. For external irradiation of mammalian cells in vitro and mammals in vivo, IDPEs of low- and high-LET radiation have been reported for radiation doses spanning in excess of three or four orders of magnitude in slightly different ranges, and for radiation dose rates both spanning over six orders of magnitude in different ranges. IDPEs of high-LET radiation in humans have been reported following internal exposure, but not external exposure. Manifestations and mechanisms of IDPEs of high-LET radiation are far less understood than those of low-LET radiation, warranting further studies that will be pivotal to assess the implications for radiation protection.

Therapeutic plasma exchange (PLEX) is an adjunctive treatment for patients with anti-neutrophil cytoplasmic antibody (ANCA)-associated vasculitis and kidney involvement. Little is known about the effect of PLEX on early changes in kidney function. This post-hoc analysis of the PEXIVAS trial investigated the effects of PLEX on changes in kidney function within 12 months. PEXIVAS was a randomized controlled trial recruiting 691 patients with ANCA-associated glomerulonephritis, of whom 349 underwent PLEX and 342 received no-PLEX. The primary outcomes of this post hoc study of PEXIVAS were change in estimated glomerular filtration rate (eGFR) from baseline and recovery of kidney function (defined as eGFR increase of 15ml/min/1.73m2 or more). Baseline eGFR was 21.7 ± 20.3 and 20.6 ± 18.7 ml/min/1.73m2 in the PLEX and no-PLEX groups, respectively. Mean improvements in eGFR at weeks two, four, and eight after initiation of therapy were greater for the PLEX vs. the no-PLEX groups. The greatest significant difference in recovery of kidney function in the PLEX compared to the no-PLEX groups was at week four (relative risk (RR): 1.41; 95% confidence interval:1.09-1.82). Increased eGFR or recovery of kidney function at week four were significantly associated with lower risk for end-stage kidney disease at week 52 (RR: 0.96: 0.95-0.97, and RR: 0.29: 0.16-0.52; respectively). Neither changes in eGFR nor recovery of kidney function differed by reduced- compared to standard-dose glucocorticoid group. Overall, our study indicates that PLEX improves early kidney function in patients with ANCA-associated glomerulonephritis.

Purpose: Growing evidence from the Japanese atomic bomb survivors, and occupationally and medically exposed groups indicates that ionizing radiation could increase the risk of various diseases of the circulatory system (DCS), even at low levels of exposure. As radiation protection systems increasingly consider the possibility of individualized radiation protection, better understanding is needed of the factors that may impact radiation-related risk, whether intrinsic (such as age, sex or genetics), or extrinsic (such as smoking). Here, we comprehensively review potential effect modification of radiation exposure and the risk of DCS in medical, occupational and environmental settings.Conclusion: Several studies indicate potential effect modification, usually detrimental, with use of anthracycline in medical settings. There was some indication that younger age at exposure increased risk for various DCS outcomes, but this was less consistent across studies and settings. Interpretation of the data is complicated by considerations of statistical power, differences in specific disease outcomes, and narrow ranges of exposure and/or potential modifiers within studies. Future studies with well-defined exposure over a wide range of ages, along with biological samples, are required to better inform the nature of these interactions.

Although leukemia in the Japanese atomic bomb survivor data has long exhibited upward curvature, until recently this appeared not to be the case for solid cancer. It has been suggested that the recently observed upward curvature in the dose response for the Japanese atomic bomb survivor solid cancer mortality data may be accounted for by flattening of the dose response in the moderate dose range (0.3–0.7 Gy). To investigate this, the latest version available of the solid cancer mortality and incidence datasets (with follow-up over the years 1950–2003 and 1958–2009 respectively) for the Life Span Study cohort of atomic bomb survivors were used to assess possible departures from linearity in the moderate dose range. Linear-spline models were fitted, also up to 6th order polynomial models in dose (higher order polynomials tended not to converge). The organ dose used for all solid cancers was weighted dose to the colon. There are modest indications of departures from linearity for the mortality data, whether using polynomial or linear-spline models. Use of the Akaike information criterion (AIC) suggests that the optimal model for the mortality data is given by a 5th order polynomial in dose. There is borderline significant (P = 0.071) indication of improvement provided by a linear-spline model in the mortality data. The low-dose extrapolation factor (LDEF), which measures the degree of overestimation of low-dose linear slope by the linear slope fitted over some specified dose range, is generally between 1.1–2.0 depending on the dose range, with upper confidence limits that sometimes exceed 10; although LDEF < 1 for the lowest dose range (<0.5 Gy), there are substantial uncertainties, with an upper confidence limit that exceeds 1.6. There are generally only modest indications of departures from linearity for the solid cancer incidence data, whether using polynomial or linear-spline models. In contrast to the mortality data, there are much weaker indications of improvement in fit provided by higher order polynomials, and only weak indications (P > 0.2) of improvement provided by linear-spline models. Nevertheless, use of AIC suggests that the optimal model for the incidence data is given by a 3rd order polynomial. LDEF evaluated over various dose ranges is generally between 1.2–1.4 with upper confidence limits that generally exceed 1.6; although LDEF < 1 for the lowest dose range (<0.5 Gy), there are substantial uncertainties, with an upper confidence limit that substantially exceeds 2.0. In summary, the evidence we have presented for higher order powers than the second in the dose response is not overwhelmingly strong, and is to some extent dependent on dose range. A feature of the dose response, which is reflected in the higher-order polynomials fitted to the data, is a leveling off or even a downturn in the response at doses >2 Gy. The linear-quadratic model is very widely used for modeling of dose response, and has been widely used in radiotherapy oncology applications as part of treatment planning. There is a theoretical basis for this model, based on the two-target model, although the data used to validate this has been mainly in vitro; there may be more complicated interactions than are implied by a two-target model, but the contributions made by these, which would contribute to higher order (than quadratic) powers of dose, may not be very pronounced over moderate ranges of dose.

OBJECTIVE:Identification of those at high and low risk of disease relapse is a major unmet need in the management of patients with ANCA-associated vasculitis (AAV). Precise stratification would allow tailoring of immunosuppressive medication. We profiled the autoantibody repertoire of AAV patients in remission to identify novel autoantibodies associated with relapse risk. METHODS:Plasma samples collected from 246 AAV patients in remission were screened for novel autoantibodies using in-house generated protein arrays including 42 000 protein fragments representing 18 000 unique human proteins. Patients were categorized based on the occurrence and frequency of relapses. We modelled the association between these antibodies and relapse occurrence using descriptive and high dimensional regression approaches. RESULTS:We observed nine autoantibodies at higher frequency in samples from AAV patients experiencing multiple relapses compared with patients in long-term remission off therapy. LASSO analysis identified six autoantibodies that exhibited an association with relapse occurrence after sample collection. Antibodies targeting homeostatic iron regulator (HFE) and synaptotagmin 5 (SYT5) were identified as associated with relapse in both analyses. CONCLUSION:Through a broad protein array-based autoantibody screening, we identified two novel autoantibodies directed against HFE and SYT5 as candidate biomarkers of relapse in AAV.

Objective ANCA-associated vasculitis (AAV) is a relapsing-remitting disease, resulting in incremental tissue injury. The gold-standard relapse definition (Birmingham Vasculitis Activity Score, BVAS>0) is often missing or inaccurate in registry settings, leading to errors in ascertainment of this key outcome. We sought to create a computable phenotype (CP) to automate retrospective identification of relapse using real-world data in the research setting.Methods We studied 536 patients with AAV and >6 months follow-up recruited to the Rare Kidney Disease registry (a national longitudinal, multicentre cohort study). We followed five steps: (1) independent encounter adjudication using primary medical records to assign the ground truth, (2) selection of data elements (DEs), (3) CP development using multilevel regression modelling, (4) internal validation and (5) development of additional models to handle missingness. Cut-points were determined by maximising the F1-score. We developed a web application for CP implementation, which outputs an individualised probability of relapse.Results Development and validation datasets comprised 1209 and 377 encounters, respectively. After classifying encounters with diagnostic histopathology as relapse, we identified five key DEs; DE1: change in ANCA level, DE2: suggestive blood/urine tests, DE3: suggestive imaging, DE4: immunosuppression status, DE5: immunosuppression change. F1-score, sensitivity and specificity were 0.85 (95% CI 0.77 to 0.92), 0.89 (95% CI 0.80 to 0.99) and 0.96 (95% CI 0.93 to 0.99), respectively. Where DE5 was missing, DE2 plus either DE1/DE3 were required to match the accuracy of BVAS.Conclusions This CP accurately quantifies the individualised probability of relapse in AAV retrospectively, using objective, readily accessible registry data. This framework could be leveraged for other outcomes and relapsing diseases.

For many cancer sites low-dose risks are not known and must be extrapolated from those observed in groups exposed at much higher levels of dose. Measurement error can substantially alter the dose-response shape and hence the extrapolated risk. Even in studies with direct measurement of low-dose exposures measurement error could be substantial in relation to the size of the dose estimates and thereby distort population risk estimates. Recently, there has been considerable attention paid to methods of dealing with shared errors, which are common in many datasets, and particularly important in occupational and environmental settings. In this paper we test Bayesian model averaging (BMA) and frequentist model averaging (FMA) methods, the first of these similar to the so-called Bayesian two-dimensional Monte Carlo (2DMC) method, and both fairly recently proposed, against a very newly proposed modification of the regression calibration method, the extended regression calibration (ERC) method, which is particularly suited to studies in which there is a substantial amount of shared error, and in which there may also be curvature in the true dose response. The quasi-2DMC with BMA method performs well when a linear model is assumed, but very poorly when a linear-quadratic model is assumed, with coverage probabilities both for the linear and quadratic dose coefficients that are under 5% when the magnitude of shared Berkson error is large (50%). For the linear model the bias is generally under 10%. However, using a linear-quadratic model it produces substantially biased (by a factor of 10) estimates of both the linear and quadratic coefficients, with the linear coefficient overestimated and the quadratic coefficient underestimated. FMA performs as well as quasi-2DMC with BMA when a linear model is assumed, and generally much better with a linear-quadratic model, although the coverage probability for the quadratic coefficient is uniformly too high. However both linear and quadratic coefficients have pronounced upward bias, particularly when Berkson error is large. By comparison ERC yields coverage probabilities that are too low when shared and unshared Berkson errors are both large (50%), although otherwise it performs well, and coverage is generally better than the quasi-2DMC with BMA or FMA methods, particularly for the linear-quadratic model. The bias of the predicted relative risk at a variety of doses is generally smallest for ERC, and largest for the quasi-2DMC with BMA and FMA methods (apart from unadjusted regression), with standard regression calibration and Monte Carlo maximum likelihood exhibiting bias in predicted relative risk generally somewhat intermediate between ERC and the other two methods. In general ERC performs best in the scenarios presented, and should be the method of choice in situations where there may be substantial shared error, or suspected curvature in the dose response.