Radiation Dose Does Indeed Matter: Proof That Invalidates The Linear No-Threshold Model

TO THE EDITOR: In their Invited Perspective (1), Duncan et al. “respectfully disagree” with our commentary challenging the Biologic Effects of Ionizing Radiation (BEIR) VII report conclusions (2). In it, we demonstrate point by point and without speculation that the BEIR committee’s conclusions are contradicted even by their own selected evidence. Choosing to ignore the evidence presented, Duncan et al., emphasizing facts that we show to be irrelevant, proclaim their unwavering belief in the correctness of BEIR VII’s conclusion that the linear no-threshold (LNT) model is valid. Since BEIR VII is a frequently cited source on the legitimacy of the LNT model, a solution to this controversy is crucial.

Duncan et al. repeat arguments made in their previous letter that we have already refuted (3). They ignore our refutations that demonstrate the need for reassessment of BEIR VII. In this brief response, we focus on 2 misconceptions [emphasis ours]:

1. “…a threshold [for cancer causation] requires processes that leave no cells harboring DNA mutations.”

2. LNT “remains the best, and certainly the most conservative, means of estimating the risk of exposing humans to varied levels of ionizing radiation.”

The existence of a threshold for radiation exposure does not require that all cells with mutations be completely repaired or removed, leaving no cells with mutations. All that is required is that fewer such cells be left with mutations after radiation exposure than before, once sufficient time is allowed for repair and removal processes to take place—usually less than 24 h. This decrease in the baseline mutation rate is the essence of hormesis.

Duncan et al. grant that endogenous processes cause mutations whether radiation—beyond the omnipresent natural background radiation—is present or not. Mutations occur continually throughout our bodies, so the baseline from which radiation operates is not zero mutations, yet some 60% of us never develop clinical cancer. This must indicate there are processes that repair or remove cells in which DNA damage could theoretically lead to cancer, a fact that is demonstrated by hundreds of studies (4,5). Thus, unrepaired and misrepaired mutations, along with double-strand breaks that exist in the absence of or after low-dose exposure, are not sufficient for the development of clinical cancer.

The claim of colinearity across dose ranges characterized by different biologic responses dissociates mathematics from its putative referent in reality. The well-established linearity of the dose response to higher-dose acute exposures (>100 mSv), as noted by Duncan et al. and undisputed by us, is irrelevant to the claim of linearity at lower doses, let alone to the existence or absence of a threshold. Furthermore, there is no credible evidence at lower doses of either linearity or absence of a threshold for carcinogenesis; both are simply assumed or based on faulty reasoning.

Although linearity may describe initial damage after low-dose exposure, the body’s complex multilevel response to this damage is nonlinear, making the overall result nonlinear. Only at high doses are these responses inhibited or overwhelmed, thereby preserving this linearity: thus, nonlinearity in the low-dose range and linearity in the high-dose range.

The second bulleted misconception above contains 2 assertions by Duncan et al.: that LNT (a) “remains the best” and (b) provides “the most conservative” estimate of risk. First, from what we have shown, LNT is not the “best,” because, while being a mathematic convenience, it is empirically false. And second, it would be the most conservative only if there were no negative side effects from overestimating risk or imputing risk where none exists and where there is actual benefit. However, as we have indicated elsewhere, the radiophobia reinforced by LNT and its corollary ALARA (as low as reasonably achievable) has several negative side effects (6). These include refusal of medically indicated radiologic imaging; misdiagnoses due to underdosing, which can lower test accuracy; and unwarranted and deadly forced relocations in the vicinity of nuclear power plant accidents.

Duncan et al. offer assumptions without evidence, irrelevant facts, and serious misconceptions, instead of evidence and rational argument. The search for the truth requires a critical reading of the literature, not uncritical acceptance of proclamations by recognized voices of authority devoid of evidence (7).

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