Challenges to the Paper “Radiation Dose Does Matter: Mechanistic Insights into DNA Damage and Repair Support the Linear No-Threshold Model of Low-Dose Radiation Health Risks”

TO THE EDITOR: The recent article by Duncan et al. (1) challenges the contentions of Siegel et al. (2) regarding the validity of the Biologic Effects of Ionizing Radiation (BEIR) VII report (3) and its underlying linear no-threshold (LNT) assumption. Duncan et al. contend that Siegel et al. fail to appreciate the appropriateness of the BEIR VII report and its LNT basis. In particular, Duncan et al. conclude:However, the linear no-threshold model remains the best, and certainly the most conservative, means of estimating the risk of exposing humans to varied levels of ionizing radiation. When considering the risks at low levels of exposure, the BEIR VII report rightfully shifted from an epidemiologic to a mechanistic approach. The BEIR VII report also appropriately considered and rejected the possibility of a threshold.

Rather than repeating the DNA arguments of Duncan et al. and Siegel et al. (1,2), this letter provides commentary that adds additional support for challenging BEIR VII (3) in general, and its underlying LNT hypothesis in particular. The following 4 arguments support the Siegel et al. contentions and further challenge the commentary of Duncan et al. and BEIR VII:

1. BEIR VII includes only a portion of the relevant dosimetric data (i.e., high-dose and dose rate data from the atomic bomb survivors and high-dose therapy patients) and excludes relevant lower dose and dose rate data. These data include (a) the extensive dosimetric documentation from nuclear power reactor and military personnel, (b) lower dose imaging data, and (c) environmental data. The BEIR VII report notes that these types of studies were evaluated, but not incorporated into the analysis. Failure to include these lower dose data provides an inherent bias and overestimates the risk of low levels of ionizing radiation.

2. BEIR VII incorporates a dose and dose rate effectiveness factor (DDREF) for low linear energy transfer data. A range of DDREF values of 1.1 to 2.3 were considered, and a value of 1.5 was deemed to be appropriate (3). The DDREF value is applied for doses below 1 Sv, and a mathematic discontinuity in the linear curve is created by reducing the slope of the dose–response curve (effects vs. dose) by a factor of the reciprocal of the DDREF below 1 Sv (3). The use of the DDREF is a tacit admission of the fallacy of the LNT approach that is a fundamental underpinning of BEIR VII. There would be no need to create an artificial DDREF factor if the LNT model were correct. Other dose cutoff values can be defined that further serve to challenge the LNT approach. For example, Siegel, Pennington, and Sacks (4) credibly demonstrate the fallacy of the LNT hypothesis as applied to medical imaging. Siegel et al. (4) note that credible evidence of imaging-related carcinogenic risk at low absorbed dose (<100 mGy) is nonexistent. A 100 mGy, 1 Sv, or discontinuity at another value adds support to challenge the credibility of the LNT approach.

3. The most recent report of the Radiation Effects Research Foundation (RERF) (5) notes a definite curvature in the data that further serves to challenge the LNT approach. RERF report 14 (5) updated the RERF report 13 (6) results and noted that formal dose-threshold analysis indicated no threshold; that is, zero dose was the best estimate of the threshold. However, Ozasa et al. note that: “Although the linear model provided the best fit in the full dose range, statistically significant upward curvature was observed when the dose range was limited to 0–2 Gy (θ = 0.81, P = 0.02) (Tables 6 and 7). The curvature over the 0–2-Gy range has become stronger over time, going from θ = 0.20 for the period 1950–1985 to 0.81 for 1950–2003, and has become significant with longer observation (Table 7).” In the preceding quote, θ is the curvature of the fit, and P is the statistical significance (likelihood test). The reader should recall that RERF report 13 (6) was a significant basis for establishing the credibility of the LNT hypothesis in the BEIR VII report (3).

4. Although the evaluation of DNA and its robust repair mechanisms are important, risk is best formulated as the integrated challenge to an organism. The effects of adaptive response, human immune system repair and mitigation, apoptosis, and other inherent protective functions also influence the final risk. Focusing solely on DNA repair is only one aspect for formulating a risk estimation model.

The BEIR VII report and Duncan et al. do not consider the aforementioned 4 factors that serve to challenge the LNT approach. As such, this letter supports the contentions of Siegel et al. (2) and encourages future BEIR reports to incorporate the challenges offered by these authors to improve future reports. In addition, the updated RERF report 14 data and low-dose and dose rate data should be incorporated into future BEIR reports to provide the best scientific assessment of the risk of ionizing radiation.

• © 2018 by the Society of Nuclear Medicine and Molecular Imaging.