Overestimation of Chernobyl consequences: biophysical aspects
After the Chernobyl accident, many publications appeared dealing with the medical consequences of the accident. In some of these, significant health effects have been reported, for example on the clean-up workers (Okeanov et al. 2004; Ivanov et al. 2008; Ivanov et al. 2009) and on thyroid cancer incidence among the population (Cardis et al. 2005; Ivanov et al. 2006; Tronko et al. 2006; Kaiser et al. 2009), while in others health effects particularly on the general population appear to be greatly overestimated. The following methodological flaws can be found in such studies: interpretation of spontaneous diseases as radiation induced, interpretation of radioactivity or dose levels without consideration of the natural background radiation, conclusions on any morbidity increase without using a proper reference population. In some studies, high figures were obviously caused by non-random case selection or inaccurate morphological assessment of the investigated specimens. Some studies are based on small sample numbers or singular observations and do not provide any statistically significant information, but create an exaggerated impression about potential consequences of the accident. Previously, we reviewed some of these studies (Jargin 2008a, b).
In one of these studies, Romanenko et al. (2006a) reported among 25 randomly selected patients with benign prostatic hyperplasia (BHP) who had resided in contaminated areas, the incidence of severe urothelial dysplasia and carcinoma in situ (CIS) to be as high as 76 and 92%, respectively. The mean 137Cs activity concentration in urine was in this cohort 6.47 Bq/l. In another publication, even higher numbers (86 and 100%, respectively) were reported in 22 random BHP patients from contaminated areas (Romanenko et al. 2006b). The doses received by these patients were unknown. The authors investigated molecular characteristics of urothelium (ubiquitination, sumoylation, upregulation of growth factor receptors, etc.) and found some differences in patients from contaminated areas when compared with the controls, which was interpreted as a consequence of long-term low-dose exposure to ionizing radiation.
Previously, this study was criticized because—among other issues—the term “long-term low-dose exposure to ionizing radiation” was inadequately used and because patients from contaminated areas and from Kiev city were combined in the investigated cohort, thus creating a basis for debate on radiation-induced malignancies in the city of Kiev (Jargin 2007a, b, c, d). These issues were not adequately addressed in the responses of Romanenko et al. (2007a, b).
It should be noted that the interpretation given by Romanenko et al. on small doses would have been applicable if the radioactivity level of concern had been counted from zero. In fact, however, the radioactivity level for most inhabitants of the contaminated areas represents just a minor increase in the overall exposure that also includes natural radiation background. For example, individual external and internal effective doses received by Kiev inhabitants during the first year after the Chernobyl accident were estimated to be about 3 mSv and 1.1 mSv, respectively, decreasing in the following years (Borovikova et al. 1991). These dose levels are comparable with global average annual doses from natural radiation background (Mould 2000). In terms of 137Cs deposition, population exposure in contaminated areas was estimated as follows: around 6,800,000 inhabitants resided on territories with a contamination level of 37–555 kBq/m2, and 270,600 inhabitants lived on territories with a contamination level above 555 kBq/m2. To the latter group belonged 115,000 evacuees, who, according to the official classification of contaminated zones, should have been resettled from areas with contamination levels above 1,480 kBq/m2 (Mould 2000). Among the evacuees from the 30 km exclusion zone, external doses were mostly below 250 mSv, but a few residents in the most contaminated areas might have received doses around 300–400 mSv (Mould 2000). Annual average effective doses received by inhabitants of the strictly controlled zones, which surround the 30 km exclusion zone, were around 40 mSv in the first year after the accident, but decreased to less than 10 mSv in the following years, thus being below the recommended upper limit for occupational exposure that is 20 mSv/y (i.e., the average over a period of 5 years, not exceeding 50 mSv in any single year) (Mould 2000). For comparison, 3,414 Uranium miners with lung cancer who worked in East Germany in the period 1946–1990 received mean cumulative exposures over 800 Working Level Months,1 which is equivalent to more than 4,000 mSv (Taeger et al. 2006). For about 13,000 miners with archived occupational data, an average exposure level of 725 WLM (3,700 mSv) was determined, including about 800 workers with the levels above 1,800 WLM (9,200 mSv) (Wesch et al. 2005; Kreuzer et al. 2006; Villeneuve et al. 2007).
Romanenko and co-workers (2007a) argued in their response that the activity concentration of 137Cs in the urine of patients from contaminated areas was on average 6.47 Bq/l and that this was probably the cause of neoplastic and pre-neoplastic lesions they observed in the bladder mucosa. For comparison, the guidance level for 137Cs in drinking water is 10 Bq/l (World Health Organization 2006). This activity concentration corresponds to recommended dose level of 0.1 mSv from 1 year’s consumption of drinking water (assumed to be 730 l). Note that natural background radiation exposures vary widely: the average is 2.4 mSv with the highest local levels being up to 10 times higher (World Health Organization 2006; UNSCEAR 2000) or even more (Ghiassi-nejad et al. 2002) without any detected increased health risks (Balaram and Mani 1994; Prekeges 2003). An increase of some 0.1 mSv due to the consumption of water with some 6 Bq/l 137Cs represents therefore a negligible addition to the natural background levels. Thus, the activity concentration of 137Cs in urine reported by Romanenko and co-workers is within the limits recommended by the WHO for drinking water. Moreover, the alkali metal cesium behaves in the organism similarly to potassium, its main excretory pathway being urine (IARC 2001). This means that the cesium concentration in urine should normally be higher than in the drinking water because of the renal concentration.
The argument that patients with benign hyperplasia of the prostate have on average more dilated urinary bladders causing “urinary retention and therefore presumably high radiation exposure to the urothelium”, used by Romanenko et al. (2007a) to explain more severe radiation damage, is hardly sustainable because most (94.5%) of beta particles emitted as a result of 137Cs radioactive decay show maximal energies of 0.51 MeV (Meyers 1987) and can, for this reason, penetrate in water for only about 2 mm (Shirley et al. 1996). Thus, the additional filling of the bladder with 137Cs-containing urine is of minor consequence for the urothelium injury by beta rays. Gamma rays with comparable energies (0.66 MeV) cause much less radiation injury in the thin urothelial layer than beta particles, because their penetration distance in tissue (half-value layer of several centimeters) is larger than that of beta particles (O’Reilly et al. 1979). Excessive filling of the bladder with 137Cs-containing urine will in fact decrease absorption of gamma- and beta rays within the urothelial layer because it gets thinner due to stretching. In any case, the reported 137Cs activity concentration of 6.47 Bq/l in the urine is much too low to cause any increase in bladder malignancy or “radiation induced chronic proliferative atypical cystitis”, which was termed Chernobyl cystitis by Romanenko and co-workers (2006a, b) and which they reported to be associated with multiple areas of severe urothelial dysplasia and CIS.
Flat CIS is most often found in association with papillary or invasive carcinoma. If left untreated, about a half or more of the CIS cases will progress to invasive carcinoma (Reuter 2004). Therefore, the high frequency of bladder CIS, reported by Romanenko and co-workers in random patients with BHP, is incompatible with the annual bladder cancer incidence in the Ukraine (50.3 cases per 100,000 inhabitants and year) reported by the same authors. The authors argued that they used “the recent WHO classification for the histological diagnosis of bladder cancer. According to this classification, CIS includes CIS from the previously classification and severe dysplasia. Therefore, the incidence of CIS may be high compared to previous WHO classification data” (Romanenko et al. 2007a). Note that urothelial dysplasia can overlap with cytological abnormalities seen in reactive conditions, which could have contributed to the reported high incidence of the CIS. The terms dysplasia and CIS should be limited to those cases where it is certain that the cytological abnormalities are neoplastic indeed. Reactive atypias and atypias of unknown significance are best reported as such. This concept was reflected in the 1998 WHO/ISUP classification and ratified in the 2003 WHO classification of bladder tumors (Reuter 2004). Another thinkable cause of the reported high CIS incidence could have been non-random selection of specimens.
Last, the Linear Non-Threshold Theory (LNT) of radiation carcinogenesis provided the theoretical basis for the overestimation of Chernobyl consequences. According to the LNT concept, a linear dose-effect correlation, proven to some extent at higher doses, can be extrapolated down to very low doses (see discussion in e.g., Friedl and Rühm 2006; Tubiana et al. 2006; Brenner and Sachs 2006; Breckow 2006). The LNT is corroborated by the following arguments: effects of ionizing radiation are of a stochastic nature; the more high-energy particles or photons hit a cell nucleus, the more DNA damage will occur and the higher the risk of malignant transformations will be (Brenner et al. 2003). This concept does not take into account that DNA damage and repair are persistent processes, normally being in dynamic equilibrium. Any living organism will probably be best adapted by natural selection to those radiation levels that occur naturally. This is obviously the case for other environmental factors such as light and ultraviolet radiation, temperature, atmospheric pressure, etc., where deviation in either direction from the optimum is harmful. For ionizing radiation this concept is confirmed by experimental and epidemiological evidence in favor of hormesis (beneficial effect of low-level exposure) (Cohen 2006; Prekeges 2003; Sanders and Scott 2006), as well as by the lacking increase of radiation-induced abnormalities in areas with elevated natural background radiation (Ghiassi-nejad et al. 2002; Balaram and Mani 1994). Natural selection is a slow process: adaptation to a changing environmental factor must lag behind its current value. Therefore, actual adaptation must correspond to some average of previous levels. Natural background radiation has probably decreased during the last millions of years, mainly due to the decay of radionuclides on the Earth’s surface and oxygen accumulation in the atmosphere due to photosynthesis, resulting in formation of the ozone layer, protecting against UV light and partly against Roentgen radiation. Moreover, accumulation of oxygen with its relatively high molecular weight has probably caused more effective absorption of cosmic radiation. There are also other reasons indicating that natural background radiation was decreasing during life existence on the Earth: declining volcanic activity bringing less radionuclides to the surface of the Earth; changing orientation of the Earth’s magnetic field and its magnetic poles and, correspondingly, varying intensity maximums of cosmic radiation which were located farther from the geographical poles in the past thus affecting more living organisms; temporarily decreasing magnetic field of the Earth during its reversals with correspondingly increasing intensity of cosmic radiation; a generally decreasing terrestrial and solar radiation, etc. Accordingly, living organisms must have been adapted to a higher background radiation level than that existing today. Therefore, the LNT concept may not be applicable to radiation doses comparable to those received from the natural background. Understanding of theses facts is required to contribute to a more realistic understanding of the anthropogenic increase of radiation and its medical implications.