Abstract
Hormesis can be explained by evolutionary adaptation to the current level of a factor present in the natural environment or to some average from the past. This pertains also to ionizing radiation as the natural background has been decreasing during the time of the life existence. DNA damage and repair are normally in a dynamic balance. The conservative nature of the DNA repair suggests that cells may have retained some capability to repair damage from higher radiation levels than that existing today. According to this concept, the harm caused by radioactive contamination would tend to zero with a dose rate tending to a wide range level of the natural radiation background. Existing evidence in favor of hormesis is substantial, experimental data being partly at variance with results of epidemiological studies. Potential bias, systematic errors, and motives to exaggerate risks from low-dose low-rate ionizing radiation are discussed here. In conclusion, current radiation safety norms are exceedingly restrictive and should be revised on the basis of scientific evidence. Elevation of the limits must be accompanied by measures guaranteeing their observance.
Background
This is an update and continuation of the previously published article, 1 which overviewed literature on medico-biological effects of low-dose radiation, concluded that current radiation safety norms are exceedingly restrictive and should be revised to become more realistic and workable. This update emphasizes bias associated with responses to radiation releases of the Soviet era, which contributed to flawed data interpretations and policy implementations perpetuating the use of the linear no-threshold theory (LNT) as the basis of radiation safety regulations. Current radiation safety norms are based on the LNT: extrapolations of a dose–response relationship down to low doses, where such relationships are unproven and can become inverse due to hormesis. 2 –6 According to the current regulations, an equivalent effective dose to individual members of the public should not exceed 1 mSv/year. The limits of effective dose for exposed workers are 100 mSv (millisievert) in a consecutive 5-year period, with a maximum of 50 mSv in any single year. For comparison, worldwide annual exposures to natural radiation sources are generally expected to be in the range of 1–10 mSv; the estimated global average is 2.4 mSv. 1,7
Recent assessments of the data on survivors of atomic explosions in Hiroshima and Nagasaki (A-bomb survivors) do not support the LNT being consistent with hormesis. 8 For solid cancers and leukemia, significant dose–response relationships were found among A-bomb survivors exposed to ≤500 mSv but not ≤200 mSv. 7,9 –11 The artificial neural network methods, applied to the data on A-bomb survivors, indicated the presence of thresholds around 200 mSv varying with organs. 12,13 The value of 200 mSv has been mentioned in some reviews as a level, below which cancer risk elevation is unproven. 11,14 According to the UNSCEAR, a significant elevation in cancer risk was observed at doses ≥100–200 mGy (milligray). 15 Among others, underestimation of practical thresholds may result from biased epidemiological research. The author agrees with Prof. Mark P Little that potentially biased studies and those of questionable reliability “should therefore probably not be used for epidemiologic analysis, in particular for the Russian worker studies considered here.” 16 –20 This recommendation may be extended onto some other studies discussed below. Moreover, the UNSCEAR evaluation of the low-dose radiation data seems to be prone to a biased data presentation tending, for example, to overestimate Chernobyl consequences; more details are in the next section and in the review. 21 Today, when the literature is so abundant, research quality, bias, and conflicts of interest must be taken into account defining inclusion criteria of studies into reviews.
Chernobyl disaster
Using the LNT, Chernobyl accident (CA) was predicted to result in a considerable increase in radiation-induced cancer. In fact, there has been no cancer increase widely believed to be a consequence of the radiation exposure except for thyroid carcinoma in people exposed at a young age. 22 –24 Although appearance of radiogenic thyroid cancers after the CA cannot be excluded, their numbers have been largely overestimated due to the following mechanisms. Prior to CA, the registered incidence of pediatric thyroid malignancy was lower in the former Soviet Union (SU) compared to other developed countries, apparently due to differences in diagnostic quality and coverage of the population by medical examinations. 25,26 Intensive screening in the contaminated territories after the CA detected not only small tumors but also advanced neglected ones accumulated in the population, misclassified after the CA as aggressive radiogenic cancers. Moreover, there was a pressure to be registered as Chernobyl victims to get access to benefits and health provisions. 27 It can be reasonably assumed that, considering corruption in the former SU, especially during the 1990s, some patients from non-contaminated areas were registered as Chernobyl victims on the basis of wrong information. There was no regular screening outside the contaminated areas, so that “imported” cases must have been averagely more advanced. These phenomena were confirmed by the fact that the “first wave” thyroid cancers after CA were averagely larger and less differentiated than those diagnosed after 10 years and later, 28,29 when the pool of neglected cancers was gradually exhausted by the screening and reliability of the registration was improved. Admixture of old neglected cases explains the fact that Chernobyl-associated thyroid cancers tended to behave in an aggressive fashion. 30 Further details and references are in the commentary. 26
The misclassification of neglected advanced cases as aggressive radiogenic cancers gave rise to the concept that the tumors supposed to be radiogenic, at least those of the “first wave” after CA, were more aggressive than sporadic ones. 28,31 –33 This had consequences for practice: Although attitudes varied, the surgical treatment of supposedly radiogenic cases was recommended to be “more radical.” 34 After 1998–1999, the surgery in some institutions switched to a more aggressive approach. 33,35 The following was recommended for Chernobyl-related pediatric thyroid carcinoma: “Radical thyroid surgery including total thyroidectomy combined with neck dissection followed by radioiodine ablation” 36 or external radiotherapy (40 Gy). 37 Some experts regarded subtotal thyroidectomy to be “oncologically not justified” and advocated total thyroidectomy with prophylactic neck dissection. 34,38 –40 Lesser resections were regarded to be “only acceptable in exceptional cases of very small solitary intrathyroidal carcinomas without evidence of neck lymph node involvement on surgical revision.” 35 It was written in a recent instructive publication that bilateral neck dissection must be performed in all cases independently of the tumor size, histology, and lymph node status. 41 This approach is at variance with a more conservative treatment of papillary thyroid carcinoma applied also in the settings of a nuclear accident. 42 The sources 43,44 were misquoted in reference 40 advocating total thyroidectomy with bilateral neck dissection for all cases of pediatric thyroid cancer. The sources 45 –47 were cited in support of the statement: “The most prevailing opinion calls for total thyroidectomy regardless of tumor size and histopathology.” 35 In fact, subtotal thyroidectomy was used or recommended in these articles, in some of them along with the total one. 45 –47 Note that many thyroid patients were girls potentially concerned with the cosmetic aspect. Moreover, total thyroidectomy with neck dissection is associated with complications such as hypoparathyroidism and recurrent laryngeal nerve palsy. 38,48 –50 The overtreatment of supposedly radiogenic urinary bladder lesions after the CA was discussed elsewhere. 21,51
Mechanisms of false-positivity have been discussed previously; among others, the misinterpretation of nuclear pleomorphism as a malignancy criterion of thyroid nodules was not unusual in the former SU of the 1990s. 51 If a thyroid nodule is found by the screening, a fine-needle aspiration is usually performed. Thyroid cytology is accompanied by some percentage of inconclusive results, when histological examination is indicated. In the former SU of the 1990s, this percentage was relatively high due to insufficient experience with pediatric material, suboptimal quality of specimens, and shortage of modern literature. The surgical specimen was sent to a pathologist, who could be sometimes prone, after the in toto removal of the nodule, to confirm malignancy even in case of uncertainty. 51 Frozen sections were sometimes used, being suboptimal for histological diagnostics of thyroid nodules. Radiophobia and cancerophobia contributed to the overdiagnosis of cancer, which can be illustrated by the following citation (from Russian): “Practically all nodular thyroid lesions, independently of their size, were regarded at that time in children as potentially malignant tumors, requiring an urgent surgical operation.” 25 The number of found nodules was additionally increased due to the iodine deficiency with enhanced incidence of goiter in the contaminated territories providing more opportunities for the false-positive conclusions; more details are in the study by Jargin. 51
The facts discussed in this section seem to be camouflaged in UNSCEAR reports. As mentioned above, the registered incidence of thyroid cancer in children and adolescents prior to CA had been lower in the former SU than in other developed countries, that is, there was a pool of neglected cases.
21
This is not clearly perceptible from UNSCEAR reports because the increased incidence 4–5 years after the accident was compared not with the pre-accident data but with that from the first years after CA, when the registered incidence already started to increase.
24
Another example is that the number of registered thyroid cancers in Ukraine prior to CA as per UNSCEAR 2008 Report
24
is higher than the corresponding data presented by IARC
52
: 39 cases for the period 1982–1985 versus 25 cases for 1981–1985. These higher figures were published with references to “communications to the UNSCEAR Secretariat”
24
and the paper.
53
However, this latter article could be found neither in online databases, nor on the website of the
East Urals radioactive trace
A tendency to exaggerate causal relationships between radiation and some diseases in the Techa river and Mayak facility cohorts, usually discussed in the context of the East Urals radioactive trace (EURT), has been noticed recently. 55 In earlier papers, no increase in cancer incidence was reported at doses ≤520 mSv or among all studied workers of Mayak facility. Existence of a threshold was held possible, although its numerical value was not specified. 56 –58 It was noticed that excessive absolute risk of leukemia had been 3.5 times lower in the Techa river cohort than among A-bomb survivors, that is, risk from the acute exposure was higher than that from protracted ones. 59,60 Later, however, the same researchers pointed out a similar level of cancer risk in the EURT cohorts and A-bomb survivors. 61 –63 An unofficial directive could have been behind this metamorphosis; potential motives are discussed in the next section. Moreover, increased risks of nonmalignant diseases—cardiovascular, respiratory, digestive—have been reported by the same and other scientists. 16 –19,64 –73 For example, the incidence of cerebrovascular disease was significantly elevated in Mayak workers with a total external dose ≥0.1 Gy, protracted over years. 74 This is indicative of a bias, in particular, of dose-dependent self-selection, noticed also by other researchers in radiation-exposed cohorts. 75,76 Individuals with higher dose estimates were probably on average more interested in medical examinations. In the health-care system of the former SU, an extent of a medical checkup has often depended on a patient’s initiative. According to a personal communication with one of the leading EURT experts Dr. Ludmila Krestinina (2014), members of EURT cohorts were preoccupied with monetary compensations. It can be reasonably assumed that persons with higher dose estimates or those residing in more contaminated areas were more insistent at medical examinations, visited medical institutions more frequently, being at the same time given more attention. This mechanism must have contributed to a more efficient detection of latent diseases in patients with higher dose estimates.
Besides, a recall bias can cause a systematic error in case–control studies: Cases would recollect facts related to the exposure better than controls, thus contributing to overestimation of doses among the cases. The UNSCEAR review could not draw any conclusions about direct causal relationships between doses ≤1–2 Gy and excess incidence of cardiovascular and other nonmalignant diseases, while mechanisms are unclear. 15 The latter figure seems to be an underestimation due to systematic errors in the epidemiological research. There is some cardiovascular risk associated with high-dose high-rate exposures; for example, patients treated by radiotherapy at doses ≥40 Gy to parts of the heart may develop heart disease later in life; some authors also discuss lower doses, 77 –79 which are still much higher than averages for the Chernobyl and EURT cohorts. The doses associated with cardiovascular damage in animals have also been higher than in the abovenamed cohorts. 80,81 The mean total dose to male Mayak workers in a study reporting an increase in cerebrovascular diseases was 0.91 Gy, protracted over years 66 ; over 90% of the Techa river cohort in a study of circulatory conditions received doses ≤0.1 Gy. 72 A relationship of atherosclerosis and cerebrovascular diseases with radiation was reported in Mayak workers exposed to external irradiation at total doses ≥0.5 Gy. 64,70 The excess relative risk for cerebrovascular diseases in Mayak facility workers was even higher than that in A-bomb survivors, 66 where the self-selection bias could have occurred as well. It is known that a correlation does not necessarily prove causality, can be caused by bias or irrelevant factors. As discussed above, the cause–effect relationships for noncancer outcomes for the low dose levels are improbable a priori. Demonstration of relationships between low-dose low-rate exposures and nonneoplastic diseases 16 –19,64 –73 cast doubt on the analogous relationships with cancer found in epidemiological studies by the same and other researchers using similar methods. 62,82 –87
Biased research about low-dose radiation: Cui prodest?
Apart from the dose-dependent self-selection and recall bias, other flaws have been identified in many papers claiming increased cancer risk from low-dose radiation, 8 for example, unfounded classification of spontaneous diseases as radiation-induced, 88,89 analyses of doses disregarding natural radiation background, conclusions about increase in incidence without correct comparisons with the control, “dose lagging, odds averaging over wide dose ranges when evaluating odds ratios, and forcing a positive slope to the relative risk dose-response curve” 90 as well as data trimming. 91,92 For an inside observer, it is evident that behind numerous publications from the former SU, overestimating Chernobyl consequences (some of them referenced in the study by Yablokov et al. 93 and commented in the study by Jargin 94 ) was a directive, which was not unusual for Soviet science. “Expected results” were discussed at scientific councils (Uchenyi Soviet) sometimes being, in fact, prescribed in advance. Desired results and conclusions could be imposed by a superior, which was favored by the authoritative management style, ingrained also in science and medicine. Since the 1980s, many former functionaries or their protégés were introduced into educational and scientific institutions of the former SU. Some of them applied pressure not only to facilitate their own careers but also to push through prescribed concepts in the guise of scientific evidence. 95,96 The author interviewed pathologists, cytologists, and other experts, who participated in diagnostics of the post-Chernobyl tumors. Many of them agreed that consequences of CA had been overestimated; and the role of vested interests was pointed out. It was also stated that sets of histological specimens from a single patient were sometimes subdivided into several ones, creating “dead souls” that influenced statistics. Causes of the overestimation of Chernobyl consequences include unreliability of Chernobyl-related research, originating from the former SU, a nonchalant attitude toward scientific misconduct in general and trimming of statistics in particular. 91,92 Other potential confounders and uncertainties were discussed in references. 13,97
According to the author’s observations, the unofficial directive to exaggerate Chernobyl consequences was issued during the Soviet era, when studies of that kind were started or planned. 89,98 –102 As mentioned above, the EURT interpretation changed later, shortly after the year 2000. This latter metamorphosis was apparently aimed at strangulation of the nuclear energy, that is, boosting of the fossil fuel prices. Considering bias associated with the evaluation of Chernobyl and EURT data sets, a reexamination of these data coupled with more realistic assessment of clinical outcomes could prove useful in further examining the dose–response relationships for low-dose radiation exposures. Among others, approaches to the revaluation could include (1) DNA examination of tumor specimens accumulated in tissue banks in order to detect “dead souls,” that is, specimens from one patient subdivided as if they were from different ones; (2) retrospective detection of nonexposed patients falsely registered as Chernobyl victims; and (3) self-criticism by researchers in order to reveal biased presentation or analysis of data. Note that politically or economically motivated misinterpretations have resulted in consequences that likely had greater adverse health effects than radiation, for example, excessive thyroid or urinary bladder interventions. 21,51 Moreover, groundless discussion of congenital malformations with reprinting in “scientific” publications of newspaper images of disfigured children 93 contributed to anxiety in pregnant women and enhanced abortion rate with interruption of wanted pregnancies after CA. 103 Considerable psychological and social impact of post-Chernobyl radiophobia is well known. 27,104
As a systematic error or bias is surmised, a question cui prodest (to whose profit) should be tackled. In the former SU, among the motives of the exaggeration of Chernobyl topic were financing, international help, publication pressure, and writing of numerous theses and articles for scientific careers. Moreover, CA has been exploited to strangle nuclear energy, 105 thus boosting fossil fuel prices. In more developed countries, antinuclear resentments have been supported by Green activists, well in agreement with the interests of fossil fuel producers and certain governments. 26 Today, however, there are no alternatives to nuclear energy: In the long run, nonrenewable fossil fuels will become more expensive, contributing to the excessive population growth in oil-producing regions and poverty elsewhere.
Hormesis and radiation safety norms
Hormesis describes processes, where a cell or organism exhibits a biphasic response to increasing doses of a substance or condition; typically, low-dose exposures induce a beneficial response, while higher doses cause toxicity. 6 Among hormetic factors are various substances and chemical elements, vitamins, light, ultraviolet, ionizing radiation, and products of water radiolysis, as well as different kinds of stress. 106,107 For factors that are present in the natural environment, hormesis can be explained by an adaptation to a current environmental level or some average from the past. This pertains also to ionizing radiation. The LNT is based on the concept that cells are altered by ionizing radiation: The more tracks pass through cell nuclei, the higher would be the risk of malignant transformation. This concept does not take into account that DNA damage and repair are normally in a dynamic balance. Natural background radiation has been decreasing over the time of the life existence on the Earth. 108 The conservative nature of the DNA repair suggests that cells may have retained some capability to repair damage from higher radiation levels than those existing today. 108 The evolutionary adaptation to ionizing radiation was explained by increased synthesis of DNA repair enzymes, activated endogenous radioprotective mechanisms, achieved, for example, by accumulation of sulfhydryl compounds and antimutagens, as well as an increase of the reserve of off-cycle cells. 109
Existing evidence in favor of hormesis is substantial, 90,110,111 which means that experimental data are partly at variance with results of epidemiological studies. Among others, there is evidence in favor of hormetic effects of low-dose radiation such as activation of DNA repair and protective apoptosis, suppression of inflammation and protection from inflammatory diseases, and stimulation of anticancer and other immunity. 112 There is experimental evidence that low-dose exposure slows aging and prolongs life. 112 Admittedly, not all experiments supported hormesis, for example, showing life lengthening of exposed mice. 113 Other studies did report life lengthening under similar conditions. 114 In animals, doses associated with carcinogenesis have been higher than those in the Chernobyl and EURT cohorts, amounting to hundreds or thousands of mGy. 7,115 –117 It should be mentioned that radiation hormesis was demonstrated also for synergistic interactions. For example, residential radon and some professional exposures may protect against lung cancer in smokers; in the Mayak facility cohort, radiation hormesis apparently protected not only against spontaneous lung cancer but also against that associated with cigarette smoking. 118 In vitro, eukaryotic cells show adaptive responses enhancing their radioresistance after a low-dose priming irradiation 3,4,119 –121 ; the mechanisms are outside the scope of this review.
For such ancient biological phenomena as hormesis and DNA repair, the data may be generalized across species, 110,111 especially if various animal species are used. Further research could quantify radiosensitivity of different species, thus enabling more precise extrapolations to humans. 122 Outstanding data, for example, “above doses of 50–100 mSv (protracted exposure) or 10–50 mSv (acute exposure), direct epidemiological evidence from human populations demonstrates that exposure to ionizing radiation increases the risk of some cancers” 123 or ×4 increase in the incidence of thyroid cancer and ×2 of benign tumors, linked to a thyroid dose of 90 mGy in children, 124 should be verified by experiments.
The benefit from a moderate exposure to ionizing radiation was reported in A-bomb survivors, 125 although these data might not be free from bias due to a better monitoring of the survivors. Occupational exposures were reported to be associated with better health, 3,4 which at least in part can be explained by the healthy worker effect. 4 Cancer mortality was found to be lower in high-elevation areas, where the natural radiation background is enhanced due to a higher intensity of cosmic radiation. 3,15,126 There are many places in the world where the dose rate from natural background radiation is 10–100 times higher than the average, for example, 260 mGy/year in Ramsar, Iran; yet no higher incidence of cancer or other radiation-related diseases have been found in such areas. 13 The screening effect and increasing attention of people to their own health may result one day in an increase of the registered cancer incidence in areas with high natural radiation background, which would prove no causal relationship. The most promising way to reliable information on low dose effects would be large-scale animal experiments. As discussed in the preceding section, integrity of all participants is needed for that.
Conclusion
Summarizing the above and previously published arguments, 1,26,51,91 the harm caused by radioactive contamination would tend to zero with a dose rate tending to a wide range level of the natural radiation background. Within a certain range, the dose–effect relationship might become inverse due to hormesis. A graph, plotted on the basis of experimental data, with a sagging of the dose–effect curve below the background cancer risk within the range of 0.1–700 mGy, is presented in the review. 116 Low doses should be analyzed separately from higher doses 127,128 to prevent unfounded LNT-based prognostications, for example, of millions of victims from nuclear accidents. 129
With regard to radiation safety regulations, a new approach is needed—to determine the threshold dose using large-scale animal experiments and establish regulations to ensure that doses are kept well below the threshold level. 8 In the author’s opinion, current radiation safety norms are exceedingly restrictive and should be revised to become more realistic and practical. Elevation of the limits must be accompanied by measures guaranteeing their observance and by openness of dosimetric data. No contraindications have been found to an elevation of the total doses to individual members of the public up to 5 mSv/year. 1 The dose rate would thus remain within the range of the natural background. Considering that development of nuclear technologies is needed to meet the global energy needs, 105 a doubling of limits for professional exposures should be considered as well. 1 Strictly observed realistic safety norms will bring more benefit for the public health than excessive restrictions that might be disregarded in conditions of disrespect for some laws and regulations. Note that disregard of written instructions was among the causes of CA. 104,130,131 The worldwide development of nuclear technologies will be possible only after a science-informed harmonization of global radiation regulatory standards and globalized control of the nuclear industry. It would facilitate construction of nuclear reactors in optimally suitable places, notwithstanding national borders, considering all sociopolitical, geographical, and geological conditions, and quality of working of local professionals, thus contributing to prevention of nuclear accidents.
Footnotes
Declaration of Conflicting Interests
The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
Funding
The author(s) received no financial support for the research, authorship, and/or publication of this article.