CSIRO - June 1994 BIOLOGICAL EFFECTS AND SAFETY OF EMR |
A study (Spalding et al 1971) that examined the effect of 800 MHz radiation on longevity in 24 RFM mice produced inconclusive results. The mice were exposed in a highly-non-uniform-field, where the average SAR was less than 1.5 W/kg, and the peak whole-body SAR in the centre of the waveguide was sufficient to result in the death of 4 exposed mice. The authors reported a non-significant increase in the life-span of the remaining 19 exposed mice (664 days) compared to the 24 controls (645 days). In a study using 50 AKR/J male mice, which spontaneously develop a high incidence of lymphatic leukaemia between 26 and 56 weeks of age, Skidmore and Baum (1974) irradiated with very short (5 ns rise time; 550 ns decay time) pulses of 447 kV/m peak electric field strength and pulsed at 5 Hz for 5 days per week for 33 weeks. The fraction of mice which had leukaemia at the end of exposure was 9/42 (21%) in the exposed group, compared with 11/24 (46%) in the sham-exposed group. However, the absence of a complete analysis of leukaemia incidence precludes any definite conclusion. The number of animals used is also small for rigorous statistical analysis. Szmigielski et al (1982, 1988) reported that exposure of mice to 2450 MHz microwaves at 2 - 3 or 6 - 8 W/kg, for 2 h per day for 6 days per week, for varying times up to 10 months; (1) accelerated the appearance of spontaneous mammary cancer, (2) decreased the time to occurrence of skin tumours induced by a carcinogen (3, 4-benzopyrene), and (3) increased the number of neoplastic nodules developing in the lungs of mice injected with cancer cells and examined 14 days later. These results suggest that RF radiation accelerated the development of three different types of tumours. In a promotion study (Szmigielski et al 1988) mice were exposed to SARs of between 2 and 8 W/kg for 2 h per day for some months following the injection of 2 x 105 sarcoma cells. The development of cancerous nodules in the lung showed a dose-dependence. However, at an exposure of 2 - 3 W/kg the effect was comparable with that which can occur from chronic non-specific stress. It has been suggested by the authors that impaired immune surveillance resulted in a lowering of resistance to neoplastic growth. The skin tumour study reported an increased number of tumours when microwave radiation followed a sub-carcinogenic dose of benzopyrene. The results of a study of 100 rats exposed for most of their lifetime at about 0.4 W/kg (Chou et al 1992) have created some disagreement in interpretation. The exposure consisted of 2.45 GHz frequency pulsed (800 pps, 10 µ pulse width) waveform modulated at 8 Hz. The whole body average SAR was estimated to range from 0.4 to 0.15 W/kg for rats weighing 200 to 800 gm, respectively. Rats were exposed for 21.5 h per day from 8 weeks of age for 25 months (i.e. lifetime). There was some evidence of effects on corticosterone level after 13 months, that was not repeatable in a smaller follow-up study. The important result was a statistically significant increase in incidence of primary malignancies in the exposed (18) compared to control (5) group. The numbers of such spontaneous cancers are too low to achieve statistical significance when classified according to each type. There is a developing opinion that a single toxic agent may be capable of producing more than one type of tumour. Comments on the protective role of melatonin, Prof. Reiter (University of Texas, Health Science Centre, San Antonio) has been quoted as saying that the suppression of melatonin by magnetic fields could result in a higher incidence of cancer in any tissue, and therefore a wide range of different tumour types. 5.3 COMMENTS ON CANCER-RELATED STUDIES
If EMR is suspected of being involved in cancer development there are a number of issues that need to be tested: (a) What cellular evidence exists, such as altered DNA, RNA, transcription rate, colony forming efficiency? (b) Is there evidence of cell membrane changes resulting in altered ionic distribution? (c) Is there evidence of immune response impairment, that may allow tumour development to proceed? (d) Is there in vivo evidence of an increased incidence of tumour growth in laboratory animals or in humans? It is virtually impossible to detect effects with epidemiology surveys without prior knowledge of the most likely biological effect. Teratologists now acknowledge that unless the full spectrum of abnormalities is known there is a very small chance of detecting even a strong teratogenic compound by epidemiology. For example, a drug such as thalidomide, which can produce a wide range of limb deformities can only be identified with certainty if all reports of every expected deformity from digital amputations to absence of one or more limbs is included. The use of a specific single effect would preclude its detection. To detect, with any degree of certainty, the effects of EM radiation, information is required on mechanisms of interaction to predict the likely consequences. The use of genetically compromised animals that are predisposed to a certain type of cancer may be inappropriate as the high base level incidence may mask small differences from the mean value thereby reducing its statistical power. Past chronic exposure animal studies have produced conflicting results, with one study (Chou et al 1992) giving either; (a) a positive or (b) a negative result depending on whether one interprets a real effect as, (i) an increased incidence in all cancers in the population, or (ii) an increased incidence of a specific cancer. The studies currently undertaken by Adey’s group represent the most thoroughly controlled chronic exposure protocol employed to date, and are being carried out at a frequency relevant to cellular telephones. Results will be available in 1995, although it will be some time later before a publication appears. Whatever the result, it will at best represent a starting point for subsequent careful study. It is most important that new studies learn from this experience. Rigorous animal experimental studies are demanding, expensive and require dedicated attention to detail for extended periods. The appropriate combination of specialised facility and trained staff is a rare commodity. Carefully worked protocols using sufficiently large exposed, sham-exposed and control groups are essential to achieve a valid evaluation of this effect. Small scale experiments comparing groups of less than 100 have low statistical power and therefore contribute little to the debate. Similarly, studies which irradiate animals with rapidly growing brain tumours are confounded by the premature death of the animal. Thus, the consequences of long term exposure to low level radiation are not tested. A recent presentation (Salford et al 1993) is such an example, where a total of 62 rates were irradiated with 915 MHz at six different exposure regimes for 7 h per day until death. Each animal had been inoculated with a large dose of glioma cells at five days prior to irradiation and died as a result of rapid growth of a large brain tumour. True scientific protocol requires the establishment of an hypothesis which must be repeatedly tested before any inference can be drawn from the results. The in vitro cell studies have provided some clues about setting such hypotheses. Perhaps the most important were the experiments of Cleary et al (1990 a) which demonstrated an altered rate of DNA synthesis and proliferation of human glioma cells after a single exposure to microwave radiation. This abnormal behaviour is consistent with early changes seen in cells that lead to tumour formation. Effects were observed at both 27 and 2450 MHz frequency and with cw or pulsed waveforms. Furthermore, he also reported the effect in cultured human glioma cells (Cleary et al 1990 b). The exposures were applied over a range of SAR, with the lowest level at which the effect was observed as 5 W/kg. Although the exposure conditions have been reported as non-thermal it is difficult to see how the exposure could avoid large thermal gradients from the cells to the cooling fluid surrounding the cell culture vessel. What makes these studies so interesting is that the effect occurs after a single 2 h exposure and lasts for up to five days. Thus, a daily exposure regimen would reinforce the effect. The connection between accelerated growth of human brain tumour cells in culture to that occurring in vivo during repeated exposure to EMR is one that deserves close examination. Hence the need for data from chronic animal studies and human epidemiology surveys. The extrapolation of results from laboratory rodents to humans is always fraught with difficulties and divergent opinions.
Relevance to Cellular TelephonesThe public concern about cellular telephones and cancer exists because of reports of an association between extremely low frequencies (ELF; 50 or 60 Hz) or police radar and cancer, and because of a lawsuit about cellular telephone use and a case of parietal lobe glioblastoma. The distinction between ELF and RF is rarely made. Reports exist of brain cancer and leukemias in epidemiology studies of nonionizing radiation at frequencies different from those emanating from cellular telephones. Little is known of human experience with cellular telephones because: (a) there are no epidemiology studies at this frequency; (b) the amount and distribution of absorption is not fully known; and (c) brain cancer is a rare event (less than 1% incidence) with a long latency period. Experimental animal studies have not been described at 915 MHz. However, a study of rats exposed to 2450 MHz radiation reported a 3-fold increase in primary malignant neoplastic lesions. Studies of different cell lines at 27 or 2450 MHz frequencies demonstrated a dose-dependent increase in growth rate that persisted for about five days, after a single 2 h exposure. Review of the literature on radiofrequency radiation and cancer yields results from eight animal studies over a 30 year period. Several of the studies reported increases of tumour incidence in the irradiated groups, especially mammary tumours and leukemias. None of these studies, used cellular telephone frequencies. National Toxicology Program (NTP) studies of chemical carcinogenesis in rodents exhibit a correlation between mammary tumours or leukemias and decreased life span. Because radiofrequency radiation doses that did not increase body temperature had no effect on life span the hypothesis has been proposed of no positive correlation between radiofrequency radiation and cancer induction. For any biological effect to become significant the body's homeostatic mechanism has to be overcome. Homeostasis uses cellular communications via molecules and ions to control the three basic functions of cells: proliferation, differentiation, and activation. Cancer promotion involves the disruption of cell-to-cell communication. One way that this can occur is through the closure of gap junctions between the cells. This disruption is both reversible and dose-responsive. If the promoter is removed, intercellular communication returns to normal. With chemical carcinogenesis, the promoting agent must be applied at a high dose, constantly, over a long period of time. During carcinogenesis a normal cell is transformed into a neoplastic cell (initiation), and the neoplastic cell grows into a neoplasm (promotion). Cell proliferation is important for both steps, and stimulation of cell proliferation would sensitise cells to the effects of endogenous DNA damage that occurs spontaneously. This is one pathway for epigenetic carcinogenesis. Thus, a carcinogenic agent does not have to alter DNA directly. An initiating agent may need only a single exposure to alter the genome and induce cancer, while promotion requires high doses, is reversible, and must continue for a long duration. Therefore, the risks from promotion are less than the risks of initiation. In general, the energy level of a cellular telephone is not sufficient to break chemical bonds. Thus, nonionizing radiation is not likely to initiate, because it cannot directly induce alterations in the genome. Except for pulsed field, cell membranes absorb much of the radiation and are a target for nonionizing radiation, and for chemical agents that act as cancer promoters. Effects at the cell membrane are consistent with promotion. Because nonionizing radiation can induce ornithine decarboxylase, the mechanism of carcinogenesis would be more likely to be epigenetic.
ConclusionCurrently there are neither data that cellular telephone use induces cancer in humans, nor any data that link nonionizing radiation from cellular telephones to tumour induction in animals. The significance of some studies showing tumour promotion is uncertain.
RecommendationsIf we consider the possible promoting effects of EMR on the development of cancer then the most relevant studies would be those carried out in animals exposed to low level radiation for extended durations. To be reliable, these studies need to use a sufficiently large population to be statistically powerful and must be designed in such a way that all known confounding variables are controlled. A major difficulty in conducting long-term exposure experiments is the problem of ensuring properly controlled exposures to large numbers of animals. It is essential to provide both an acceptable, low stress, living environment for the animals and a uniform microwave exposure for the entire population. RF coupling to the animals body will differ as a function of number of animals in the group, distance between them, and orientation of each animal in the RF field (D’Andrea & de Lorge 1990). At the same time, there is good reason to have concurrent exposure to normal environmental pollutants that might potentiate the effects. Finally, the protocol needs to be robust and repeatable to allow exact duplication of the study in an independent laboratory. It is well known that the distribution of RFR in an exposed object depends on many factors including frequency, orientation of exposure, dielectric constant of the constituent tissue. The design of experimental protocols is critical if the results are to provide meaningful extrapolation to a particular RF source. Cellular telephones are used in a specific manner. Most people would hold a phone to the same ear in the same orientation and proximity to the skull. Usually one would expect the antenna to be close to the parietal bone (although many airport officials have a peculiar habit of holding the large portable phones in front of their mouth so that they look across the top of the antenna). However, assuming normal usage patterns it would make sense to design experiments so that the RF source was located towards the lateral aspect of the skull. Chou et al (1985a) found significant differences in local SARs in eight different regions of the brain of rats and these all changed in each of seven different exposure arrangements. Lai et al (1984a) reported a difference in microwave response with pentobarbital depending on whether the rat was facing toward or away from the source of irradiation in a waveguide when the average whole body SAR remained constant; patterns of energy absorption in the brain differed substantially.
[CONTENTS] [GLOSSARY] [REFERENCES] Current Research Abstracts Index
Please e-mail comments, information and updates to DON MAISCH: |