Radiological toxicity of DU  K. BAVERSTOCK, C. MOTHERSILL & M. THORNE
(Repressed WHO Document) 5nov01

 Keith Baverstock World Health Organization
European Centre for Environment and Health
Hermann Ehlers Strasse 10
D-53113 Bonn, Germany
Tel: +49/228 – 2094 430 Fax: +49/228 – 2094 201

Carmel Mothersill Dublin Institute of Technology,
Kevin Street, Dublin8, Ireland
Tel. +353-1-4027509, Fax. +353-1-4023393

Mike Thorne Mike Thorne and Associates Limited
Abbotsleigh, Kebroyd Mount, Ripponden, Halifax,
West Yorkshire, HX6 3JA, UK
Tel/Fax: +44-01422825890


Background: The military use of depleted uranium (DU) and/or recycled uranium (RU) has given rise to public concern as to the impact on public health of exposure to environmental sources. Exposure to soluble natural uranium, through drinking water and the food chain, is ubiquitous. After military use, DU / RU are present in the environment either as metal or as oxide dusts. Due to the low specific activity of uranium, the potential effects of exposure are generally attributed to chemical toxicity. Insoluble particulates may be an exception.

Results: DU/RU dusts are a mixture of oxides of differing solubility, such that, if retained in the lung, partial dissolution occurs over the time scale of about a month. As DU has been shown to be capable of transforming human cells to a tumourigenic phenotype without the involvement of radiation, such particles present a unique radiological/chemical toxic hazard. The bystander effect may be of relevance where an alpha-particle emitter of low specific activity is distributed over the lung.

Conclusions: The health risks of exposure to DU/RU are likely to be only partially reflected by the radiation dose per received. Further work on the chemical transforming ability of DU, the potential for an interaction between its chemical and radiological toxicities and the significance of the bystander effect in this context is required to fully estimate the public health significance of exposure to DU/RU.

[1] Disclaimer

The ideas and views expressed herein are those of the author and should not be taken to necessarily represent those of the World Health Organization.

1. 0 Introduction

The military use of depleted, and or reprocessed uranium, in Iraq and the Balkans, as penetrators in various munitions and as armour, has raised questions as to the radiological toxicity of these forms of uranium. Although it should be emphasized that there is no established evidence (as opposed to media claims) that links exposure to the environmental residuum of these weapons to diseases that would normally be associated with radiation, that populations live close to contaminated zones inevitably gives rise public health concerns. In addition, claims of illness in military personnel who have served in theatres where DU has been employed are currently being investigated. In this connection the UK Royal Society (RS 2001) have examined the health hazards of DU munitions to military personnel and the United Nations Environmental Programme has carried out an environmental assessment. (UNEP 2001)

This paper is concerned with the health implications of exposure to DU after its military use. Although the primary emphasis is with its radiological toxicity, aspects of chemical toxicity are also addressed.

Various studies on employees in the Uranium processing industry (eg. Ritz 1999; Archer 1981; Cardis and Richardson 2000; Dupree, Cragle et al. 1987; Checkoway, Pearce et al. 1988; Kathren and Moore 1986, Kathren, McInroy et al. 1989; Loomis and Wolf 1996; McGeoghegan and Binks 2000; Ritz, Morgenstern et al. 2000) do not present a clear picture of the health effects of exposure to uranium due to small numbers and potentially confounding exposures. However, associations with lymphopoietic, lung, bone and kidney malignancies cannot be ruled out. At the same time, uranium is also ubiquitous in the natural environment. It is often argued that this natural exposure can be used as a "benchmark" for exposures such as that to DU after its military use. We show here that this is not necessarily the case, and that both the chemical form and the route of entry into the body may have a critical influence on toxicity.

Following military use, DU will be distributed in the environment either as the metal, in anything from whole armaments to fragments and shards, or as oxide particulates with diameters ranging from the order of microns to nanometres. The dissolution of the metal into aqueous solution will be a slow process, leading to the contamination of groundwater and soils over a period of several hundred years. Uptake by plants from contaminated soils will be limited, as uranium is relatively strongly excluded from root uptake (Sheppard and Evenden 1988). Overall, the natural uranium content of soils, plants, animals and drinking water will be somewhat increased over the area in which the depleted uranium is dispersed. In these circumstances, the chemical toxicity of the additional uranium is of much greater interest than its radiological toxicity. Furthermore, chemical toxicity will only be of importance if the depleted uranium is present at concentrations that are comparable to, or higher than, those of available natural uranium (i.e. excluding that component of natural uranium that is incorporated in uraniferous minerals and hence is not available for uptake). In most soils this concentration is a few parts per million. (WHO 2001)

 1.1 The origins of depleted uranium and its military application

Uranium is a naturally occurring element with isotopes of long radioactive half life and, therefore, low specific activity. The principal isotopes in natural uranium are 238U, 235U and 234U. Depleted uranium (DU) is a waste product of non-nuclear enrichment processes (e.g., gaseous diffusion of uranium hexafluoride) in which the content of 235U in natural uranium is enriched, leaving the DU with a reduced content of the lower atomic weight isotopes. The enriched uranium can be used to generate 239Pu by partially "burning" it in a nuclear reactor. After extraction of the 239Pu and other radioisotopes of elements other then uranium, the residual uranium can be enriched for further burning and plutonium production, generating additional uranium depleted of the lower atomic weight isotopes. As this material, which has been subject to nuclear processes, is potentially contaminated by isotopes generated by the neutron flux in the reactor (e.g. technetium, plutonium, neptunium, americium) it should be distinguished from the material arising from the first enrichment process, and here it is termed reprocessed uranium (RU).

In terms of its physical properties, uranium is a dense and hard metal that is pyrophoric. It is these properties that give the effectiveness at penetrating armour and destroying tanks and their occupants. On burning, uranium produces a dense smoke, which, in a confined space, is rapidly suffocating.

1.2 Initial considerations in estimating the toxicities of environmentally distributed DU and RU

The isotopic composition of an element makes no substantial difference to its chemical properties but may influence its radiological properties though modification of its specific activity. Since 235U and 234U have higher specific activities than 238U, the radiological toxicity of DU is expected to be lower than that of natural uranium by about 40%.

The specific activity of RU will depend on the extent to which the uranium is contaminated by fission products and other nuclides produced by the neutron flux in a nuclear reactor, and not removed by the subsequent processing.

There are only very limited animal and human data on the radiological and chemical toxicities of DU and none relating to RU, but there is much more abundant evidence from the ubiquitous exposure to natural uranium, particularly in terms of its chemical toxicity. These data can be used as a reliable guide to the effects to be expected from DU, provided account is taken of the chemical form and route of entry into the human body. Limited epidemiological data are available from studies of workers in uranium milling plants who were exposed to dusts containing uranium. Studies of the behavior of inhaled dusts in the lung have resulted in models from which the radiation doses to lung and other body tissues can be calculated. Such models provide both absorbed and equivalent doses in Gy or Sv per Bq of inhaled dust, contingent on the solubility and size distribution of the dust particles. Thus, if the specific activity (Bq/ unit mass) of the inhaled material, characterized by its solubility and particle size distribution, is known, the radiation doses to the lung and other tissues can, in theory, be estimated. (ICRP 1995).

The burning of uranium produces a mixed oxide dust, part of which is relatively soluble in lung fluids and a part of which is insoluble. As the burning of DU arises almost exclusively in military operations, reliance has to be placed on the limited data released by the military authorities. Much of this information is summarized in a US Department of Defense Report (CHPPM 2000). According to this report, DU burns on impact with a hardened target, such as the armour of a tank. The extent of burning depends upon the characteristics of the impact and factors such as the degree of fragmentation of the DU. The extent of release of DU oxides to the wider environment also depends on the particular situation. In some cases, where the DU penetrates the target, most of the DU oxides will be retained within the structure of the target. However, a hardened target may lead to fragmentation and burning of the DU in the open and a release of the DU oxide dusts to the environment.

Of relevance to environmental exposures to DU/ RU are the following:

  1. Total mass of DU/ RU delivered into the environment.
  2. Proportion of that mass that hits a "target".
  3. Proportion of the material hitting the target that burns to produce DU/RU oxide dusts.
  4. Proportion of that dust that is released to the wider environment.
  5. Mobility and lifetime of the dust in the environment.
  6. Exposure of humans to the dust and its respirability.
  7. Proportion of DU/ RU dust that is soluble in the lung.
  8. Particle size distribution of the DU/ RU oxide dust. (This is also related to solubility.)
  9. Specific activity of DU/RU oxide dust for each of the radionuclides present.

1.3 Evaluating the extent of DU/ RU oxide contamination of the environment

 In any given instance of environmental contamination by DU/ RU, the situation will need to be assessed by environmental monitoring. However, the CHPPM report gives some indications that would allow an initial "desk" assessment, from readily obtainable information, to be made. Given that the total mass used is available, the CHPPM report estimates that, for an aerial attack about 10% of penetrators hit a target. It can, therefore, be assumed that about 90% of the material will be on the ground or buried, in a metallic form. In a tank-to-tank battle the proportion of hits on targets will be greater.

The extent to which the DU hitting a target burns, and the fraction of oxide released to the environment depends on the circumstances and could be anything from a few to several tens of percent. According to CHPPM, a representative figure could be 70% burned, up to half of which is released as highly insoluble oxides. (RS 2001)

Little quantitative information exists on particle-size distribution. Generally, it is concluded that a substantial fraction falls within the respirable size range and that ultra-fine particles, which have a tendency to coalesce, are also formed. (RS 2001)

The CHPPM report has little to say on the question of RU. It notes that traces of other nuclides, notably plutonium, neptunium and americium are contained in some of the so-called DU used in armour and some munitions but that this additional activity "adds less than one percent to the internal radiation risks." However, the report leaves open the question of whether, in the case of all munitions, this 1% is a maximum.

It can, therefore, be concluded that environmental contamination by DU/ RU does have a potential for both chemical and radiological toxicity, thus creating the necessity for assessing the public health impact for those living in contaminated zones.

2.0 Exposure Routes and Biokinetics of Uranium

Because of the importance of uranium separation, enrichment and fabrication in both military and civil applications of nuclear power, there is over fifty years of experience in working with the metal and a wide variety of its chemical compounds. Over that period, tens of thousands of workers have been exposed, both by ingestion and inhalation. In consequence of this operational experience and complementary experimental studies on both humans and animals, there is comprehensive understanding of the biokinetics and toxicology of uranium. This understanding is relevant to an appreciation of the specific issues relating to the use of depleted uranium in projectiles and armour.

Uptake of ingested uranium from the gastrointestinal tract is relatively low. Even for soluble salts of the element or for uranium incorporated in food, the fractional gastrointestinal absorption (f1) is less than about 0.05. Results from a recent study on uranium in drinking water from Finland (Kurttio, Auvinen et al., in press) find a value for f1@ 0.003. This is the first human study for which this value has been determined. It is possible that some uranium in well water is in an insoluble form and that this accounts for the relatively low value of f1. For insoluble salts, such as UO2, the fractional absorption is much less, typically less than 0.01 (ICRP, 1995).

The uptake of inhaled uranium to the systemic circulation can be much greater. Typically, about 60% of inhaled material is deposited in the respiratory system, with the remainder lost upon exhalation (ICRP, 1994). For soluble salts of uranium, almost all the deposited material is transferred to the systemic circulation on a time scale of a few days. For insoluble uranium, the situation is rather different. Mechanical processes clear the majority of uranium in the upper respiratory tract, including the bronchial tree, on a time scale of hours to days. The cleared material is swallowed and is almost entirely lost by faecal excretion. However, insoluble salts of uranium deposited in the deep lung (the pulmonary parenchyma) are typically retained with a biological half life of around 100 days (or longer for high-fired UO2). Clearance of this material occurs by both mechanical clearance, often of particles ingested by phagocytes, and by solubilisation. A few percent of inhaled insoluble material reaches the systemic circulation by dissolution. A further small fraction may be translocated as particles to the tracheo-bronchial lymph nodes and from there to the systemic circulation (ICRP 1994, ICRP 1995).

Once uranium has reached the systemic circulation, its subsequent biokinetics is well described by the model developed by the ICRP (ICRP 1995) (see Figure 1).

A large fraction of uranium that enters the systemic circulation is taken up and retained in mineral bone. Smaller fractions exchange with the liver and general soft tissues. Although there is a very limited degree of excretion from the liver to the gastrointestinal tract, almost all excretion is in the urine. It is the urinary excretion component that is of specific relevance to the chemical nephro-toxicity of uranium. This urinary excretion path is illustrated schematically in Figure 2 (based on Leggett 1989).

In body fluids, the main form of uranium is thought to be the uranyl ion, UO2++ (Leggett 1989). However, in the blood plasma approximately 40% of uranium is present as transferrin complexes and 60% as low molecular weight anionic complexes. These low molecular weight anionic complexes are filtered rapidly by the glomerulus and enter the lumen of the kidney tubule. The rapidity of this process may be illustrated by noting that, in the first 24 hours after entry of uranium nitrate into the systemic circulation, around 80% will have been filtered by the glomerulus (Leggett 1989).

As the filtered uranium complexes pass along the renal tubules they are subject to a fall in pH. This results in their partial dissociation. Whereas some complexed uranium plus a proportion of the uranyl ions produced on dissociation is excreted in the urine, the remainder of the uranium binds to the luminal membranes of the renal tubules. The bound uranium is removed from the luminal membranes by combining with ligands in the urine, shedding of microvilli, sloughing of dead cells, or entering cells. The rate of loss by each of these processes is thought to be dependent on the magnitude of the exposure to uranium, such that the fraction of uranium retained in the kidneys increases with increasing administered amount (Leggett 1989).

It is thought that the mode of entry of uranium into renal tubule cells may be primarily by endocytosis. Intracellular accumulation is mainly in lysosomes, with microcrystals formed at high concentrations. Destruction of the lysosomes then releases these microcrystals into the cytosol.

Although intracellular uptake is primarily into lysosomes, smaller amounts of uranium accumulate in the nucleus, mitochondria and other intracellular organelles. (Leggett 1989)

Overall, uranium-containing debris may be retained for an extended period in the lumen of the tubule or in reticuloendothelial cells.

Retention of uranium in the kidney is known to give rise to a variety of biochemical effects that may have implications for the clinical toxicity of the element (Leggett 1989). These include the following:

At an overall tissue level, the kidney may develop tolerance to uranium exposure after repeated or chronic exposure, but this is associated with regenerated cells with a degraded brush border. Impairment of function can be associated with such tolerance. For example, tolerant animals have been observed to exhibit high urine volumes and a diminished glomerular filtration rate. It has been concluded that acquired tolerance to acute affects does not prevent chronic damage. (Leggett 1989)

Conventionally, it has been assumed that if kidney concentrations of uranium are maintained at less than 3 m g/g, symptoms of clinical toxicity will be avoided. However, this limiting concentration was based on tests of limited sensitivity and on criteria for toxicity that are less stringent than would now be employed. In view of these considerations, it has been suggested (Leggett 1989) that it may be prudent to lower this long-standing level by one order of magnitude.

 3.0 The Relative Significance of Chemical and Radiological Toxicity for Depleted Uranium

The oxide particulates may be much more refractory to dissolution than the metal, if they are primarily composed of UO2. Refractory particles inhaled at the time of their production or subsequently, as a result of resuspension, could be of greater significance radiologically than through the chemical toxicity of their uranium content. This is because such particles can be retained in various organs and tissues, including the respiratory and reticuloendothelial systems, irradiating their surroundings. If such particles are leached only slowly, they will contribute to only a limited degree to an increase of uranium concentrations in the kidneys.

The distribution and retention of inhaled radioactive refractory particulates has been studied extensively. In particular, a great deal of work has been undertaken on high-fired PuO2. Particles, with aerodynamic diameters of up to a few tens of micrometres are readily inhaled. Particles with aerodynamic diameters of more than a few micrometres are mainly deposited in the upper part of the respiratory tract (the nasal passages, trachea and larger bronchi) and are largely cleared by mechanical action on a time scale of a few hours. Smaller particles penetrate more deeply into the lungs and sub-micrometre particles are deposited mainly in the respiratory tissues (the pulmonary parenchyma) comprising the bronchioli and alveoli. (ICRP 1994)

Material deposited in the alveoli is beyond the limits of the region from which direct mechanical clearance can occur (ICRP 1994). Therefore, clearance from this region is due mainly either to solubilisation or to incorporation and transport of particles in phagocytes (the alveolar macrophages). These macrophages may either migrate to the bronchial region and be mechanically cleared, or they may penetrate the alveolar interstitium and be carried to the regional lymph nodes.

In the 1970s, there was considerable interest in whether such focal sources of radiation (‘hot particles’) were of greater concern than homogeneous irradiation of respiratory tissues to a similar average radiation dose. In general, it was found (Burkart and Linder 1987) that such focal sources were no more radiotoxic than uniform irradiation and could be substantially less toxic. The latter result was attributed to cell sterilisation effects around the focal sources, as sterilised cells are incapable of reproduction and cannot be the precursors of cancer. However, some caution should be exercised in interpreting the results that were obtained, because the work was largely based on the assumption that only cells that are ‘hit’ by radiation tracks can be transformed to neoplastic precursors. More recent studies have demonstrated a bystander effect, in which unirradiated cells close to irradiated cell populations can exhibit genetic alterations. It may, therefore, be prudent to examine again the question of whether focal sources of irradiation could induce a spectrum of effects that differs from that induced by more uniform irradiation. In the specific context of uranium, it is of interest also to consider whether the enhanced soluble uranium concentrations that could exist in the vicinity of individual particles or aggregates could interact synergistically with the localised irradiation of tissues, particularly if some of the effects of irradiation are mediated by substances released from the irradiated cells.

In considering whether such effects could occur, it is appropriate to recognise that particles could accumulate or aggregate in interstitial tissues of the lung, in pulmonary lymph nodes or in reticuloendothelial tissues. In the context of reticuloendothelial tissues, an analogy can be drawn with the colloidal radiographic contrast medium Thorotrast (ThO2). This was found to give rise to substantial aggregates in the liver, spleen and bone marrow, and excesses of both liver cancer and leukaemia have been observed in the exposed populations (Van Kaick, Muth et al. 1986). However, too much weight should not be placed on this analogy, as the masses of Thorotrast used were large (around 25 g per patient) and it was introduced directly into the systemic circulation giving enhanced opportunities for aggregation and deposition into reticuloendothelial tissues.

4.0 Heath impacts of uranium

4.1 Inhalation of uranium oxide dusts

Breathing uranium containing dusts is an established occupational hazard with which clear health consequences are associated. Most information relates to uranium miners, whose exposure to uranium ore dusts is compounded by collateral exposure to radon daughter products. The much greater activity concentrations of radon daughters in air leads to relatively larger doses to the lung than from the uranium itself, and thus the established yield of lung cancer from such exposures is attributed to radon. However, workers in uranium milling plants, where the radon daughters are not so abundant, also show indications of increased disease that could be due to radiation (Cardis and Richardson 2000). Lung cancer is elevated in a number of studies (see Cardis and Richardson 2000; Ritz 1999; Checkoway, Pearce et al. 1988; Loomis and Wolf 1996), although it should be noted that the situation is compounded by exposures other than to internal a -emitters and, in individual studies, numbers are generally small.

In the most recently reported study of uranium plant workers at Springfields in the UK (McGeoghegan and Binks 2000), where uranium ore was handled, there was a substantial healthy worker effect and no absolute excess or trend with dose for lung cancer.

In other stages of the uranium processing industry, where soluble uranium may be inhaled as aerosols, there are indications of increases in lymphopoietic (Loomis and Wolf 1996, Ritz, Morgenstern et al. 2000) brain, kidney, breast, prostate (Loomis and Wolf 1996) and upper aerodigestive tract (Ritz, Morgenstern et al. 2000) cancers.

In a response to an editorial (McDiarmid 2001) in the British Medical Journal, Alvarez has drawn attention to health effects seen among uranium process workers, as described in an unpublished report (see ). As noted, (Ritz 1999) there were positive associations for several cancer sites with chemicals used in the uranium processing industry. It is, therefore, clear that working in the uranium processing industry is associated with a number of different types of cancer, but whether this is due to insoluble or soluble uranium or other chemicals used in the processing is not clear.

The uranium dusts encountered in the milling process may be more insoluble than the dusts generated by burning DU and are almost certainly of different particle size distribution. Burning metal has the tendency to produce sub-micron particles as well as the more usual 1 to 10 micron Activity Median Aerodynamic Diameter particles that are generally associated with radiological toxicity. Such sub-micron particles present some features that may be significant in evaluating the toxicity of DU (as opposed to natural uranium). These ultra-fine particles may be more soluble in physiological fluids, thus creating a local environment of enhanced uranium concentration in the cells proximal to the particle of DU-oxide. In this respect it is notable that DU-UO2 2+ cation is capable of transforming human osteoblast cells in culture to a tumourigenic phenotype (Miller, Fuciarelli et al. 1998). Similar transformation can be achieved with nickel and, to a lesser extent, with lead, leading to the conclusion that this transformation may have little to do with the radioactivity of DU. This conclusion is confirmed by the small fraction (0.0014%) of cells hit by alpha particles at the uranium concentrations used.

It is relevant to note that nickel is an established carcinogen (IARC 1990) and has been shown to induce a genomic instability similar to that induced by radiation (Coen, Mothersill et al. 2001).

Partially soluble dust particles, either because of chemical composition or size, produce a unique situation in which a volume of tissue a few cell diameters in radius, around the particle will be subject to both a relatively high concentration of UO22+ and the occasional alpha particle from decay of the 238U. A 1m m particle of pure 238U weighs 5.8x10-6m g and on average emits 2 alpha-particles per year. Assuming that over a period of weeks half the material dissolves and is retained within a volume of radius 3 cell diameters, or 30m m, the concentration of UO22+ in this tissue volume is about 20m g/g or 0.8mM – well in excess of the 10m M concentration at which cellular transformation associated with (or leading to) tumour formation in nude mice was seen.

For a total intake of 1 mg of such a dust and assuming that 25% is retained for a long period in the lung of which 50% behaves as a Class M (ICRP 1994) material and dissolves relatively slowly, the remainder being insoluble, there would be about 0.4 x 108 such foci with 20% (8 x 106) also experiencing one alpha passage in the first month. This is not a situation that has been experienced in any exposure situation for an alpha or any other emitter in the lung. It is not possible to extrapolate the risk of such an exposure from human experience. In particular the risk to the lung of exposure to DU dusts cannot be inferred from the experience gained from uranium miners, or from survivors of Hiroshima and Nagasaki, upon which the current ICRP radiological protection standards are based.

A second factor is the potential for small particles to become trapped in the interstitial spaces where they may form aggregates. Clearance is likely to be to the local tracheobronchial lymph nodes (TBLN), where they may be retained indefinitely.

A significant excess of lymphatic and haemopoietic cancers, other than leukaemia, (4/1.02) in uranium mill workers, whose concentration of uranium in urine was elevated, is noted (Archer, Wagoner et al. 1973). It is suggested that these malignancies could have resulted from an accumulation of long-lived radioactive materials in the lymph nodes.

However, Baverstock and Thorne (Baverstock and Thorne 1989), in reviewing evidence for consequences of irradiation of the lymphatic system from material retained in the tracheobronchial lymph-nodes, concluded that, in spite of the real possibility of substantial doses, there was little reason to expect an excess of lymphatic leukaemia. They noted, however, that their arguments could not be wholly conclusive.

Furthermore, small particles (10 to 100nm) are capable of passing through the pulmonary blood vessels into the blood stream. Experience with directly injected colloidal particles of thorium oxide, in the form of the x-ray contrast medium Thorotrast, shows that such particles have a tendency to aggregate in reticuloendothelial tissues, where they are retained, if insoluble, over long periods. In the case of Thorotrast, the long-term consequences were liver cancer and leukaemia. Doses from the injection of Thorotrast are likely to have been very much larger than could be obtained from inhaling DU smoke, as the direct transfer through pulmonary blood vessels is only a minor lung clearance route.

Overall, there seems to be a compelling case for investigating whether uranium, internally incorporated through inhalation, has a combined chemical and radiological carcinogenic potential, which can potentially lead to cancers in the lung and other parts of the body, including the lymphatic system, the bone marrow, the bone and the kidney. Therefore, the extent to which DU, present in the environment as dust and smoke from burning metal, is able to cause these consequences, though a combined radiological and chemical effect, is a matter for further research.

The implications of the bystander effect also need to be considered in this context. It has been convincingly demonstrated that changes, similar to those caused directly by irradiation, can be wrought in cells growing close to a cell that has been irradiated, or even if they receive activating signals in medium harvested from irradiated cells, even though the changed cells experienced no ionising event. Such changes include genomic instability, widely associated with the cancer process, and even mutations, also widely believed to be related to cancer induction (Mothersill and Seymour 2001). The basis for this phenomenon is not well understood, but it has been demonstrated that a calcium pulse occurs and resolves within 5 minutes of exposure of non-irradiated cells to medium harvested from exposed cells. Alpha particle radiation is known to be a potent cause of bystander effects, particularly in the form of genomic instability and, since heavy metals can also cause instability (Coen, Mothersill et al. 2001), there is a strong case that the mixed radio-chemical exposure may be acting in this context.

As directly inflicted DNA damage is precluded as a cause of the bystander effect, it can be inferred that a chemical agent is transmitted from the irradiated cell and that this changes the state of the recipient cell in an apparently irreversible manner. A recent study (Belyakov, Malcolmson et al. 2001), using micronucleus formation as an endpoint and a micro-beam facility capable passing a single alpha particle through the nucleus of a specific cell, showed a three-fold increase in damaged cells within the environment of the irradiated cell. Typically, 5000 cells were scored with some 100 excess damaged cells. However, excess affected cells were found at distances of mm from the irradiated cell and thus the number of potentially affected cells per particle can be very large. Within 1 mm radius of the irradiated cells there are approximately 106 cells, thus if the same ratio of affected cells applied some 2 x 104 could be affected.

The bystander effect is predominant at low tissue doses, where few cells experience an alpha particle passage. At higher doses, recipient cells increasingly experience alpha passages themselves, with a high probability of cell killing and almost certainty of inducing other changes, thus reducing the relative effectiveness of the bystander effect. For this reason, uranium particles, which emit few alphas, would have a greater chance of inducing effects through the bystander mechanism than "hotter" particles.

The implication of the combined chemical and radiological transforming capability of uranium and the bystander effect, means that, in estimating its significance in causing cancer, the simple assumptions, based on committed effective dose, ie (committed absorbed dose to the lung, modified by a radiation weighting factor for the fact that the radiation arises from alpha particles) as has been adopted in recent reports by the Royal Society (RS 2001), the WHO (WHO 2001) and UNEP (UNEP 2001) would be an inadequate basis for predicting risks.

 4.2 Other considerations

The usual assumption, based on the specific activity of uranium, standard tissue and radiation weighting factors (ICRP 1991) and the distribution of uranium between different tissues, is that impairment of kidney function will always be more important that any carcinogenic effect. This assumption can, however, be questioned on two grounds, namely the potential for synergy between chemical and radiation toxicities, and the bystander effect, as discussed above.

In the experiments with osteoblasts (Miller, Blakely et al. 1998), the concentration of UO2++ was 10m M, which is close to the 0.3m g/g level in the kidney assumed to be below the threshold for toxic effects. In the transformation assay, this produced a ten-fold increase in the tumourigenic phenotype with about 1 in 105 cells being hit by an alpha particle. It is feasible to explain the transformation in the osteoblasts by the bystander effect alone, but the similar level of transformation brought about by the same concentration of nickel ions cannot be explained radiologically.

If there is indeed a synergistic effect between the chemical and radiological properties of uranium, why is exposure to naturally occurring uranium apparently without radiological health consequence? One answer to this question is that natural uranium is almost entirely ingested. The fraction of even soluble uranium crossing the GI tract is low (typically around 0.02, see ICRP Publication 69 (ICRP 1995)), most being excreted in faeces. In the occupational context, the primary route of entry will be inhalation of aerosols. Where the uranium is soluble, the transfer to blood of deposited material is rapid and complete (ICRP 1995). Potentially much higher body burdens could be acquired in this way.

Among the soft tissues in which systemic uranium locates are the testes. This raises the prospect of hereditary effects arising from systemic burdens. The non-specific nature of the location of uranium at the cellular and sub-cellular levels implies that all testicular cells are at some degree of risk, including the spermatogonial stem cells. The relevance of the transforming effect observed for uranium is problematic. If that transforming ability is mediated by mutations then a synergy may also be expected here. In the Miller study (Miller, Blakely et al. 1998), changes in gene expression and sister chromatid exchanges were observed, leaving the question open.

5.0 Practical public health implications of the use of DU/RU in two theatres of war, the Balkans and Iraq/Kuwait.

Ammunitions containing DU and RU have been used in the Balkans and Iraq/Kuwait. Comparing the two instances there are important differences that have a bearing on public exposure to DU/RU. (RS 2001). In the Balkans, the ammunition was exclusively fired from aircraft, whereas in Iraq the tank-to-tank battles also took place. In air-to-ground fire, fewer DU/RU rounds hit targets such as tanks, most, as much as 90 to 95%, becoming buried in the ground. Thus, only 5 to 10% was at risk of fragmentation and burning. In Iraq/Kuwait, a larger percentage will have hit hardened targets and burned to produce the oxide smoke and dust. The United Nations Environment Programme has carried out an environmental assessment in Kosovo (UNEP 2001).

Metallic DU/RU buried in the ground will slowly dissolve (over centuries) so somewhat enhancing the natural level of uranium in the natural environment. It is legitimate to place the risks of this exposure in the context of naturally occurring uranium levels in the environment and it seems unlikely that the small increase in uranium levels this will entail (except in the circumstance that a penetrator lodges in very close proximity to a drinking water well) will constitute a hazard to health. Given the climatic conditions in the Balkans, it seems unlikely that re-suspension of the dusts resulting from the 5 to 10% of munitions burning will lead to prolonged exposure of the population by this route although in the first year or two hot summer weather may have led to some resuspension. In any case weathering and leaching of the dust on the ground will result in a lowering of its potential toxicity. The health risks to the civilian populations, peacekeeping troops and aid workers in Balkans are, therefore, likely to be minimal in the future, the principal risks being confined to those who were on the ground during the actual time of use of the weapons, namely a small minority of the indigenous population and the Serbian troops.

The situation in the Iraq/Kuwait theatre, for which there is no environmental assessment, is somewhat different. Given the higher percentage of burned DU/RU in the tank-to-tank fire, the generally dry and arid climatic conditions of the area and the presence of a civilian population at the time of the battles, the potential for exposure to dusts and smoke of the combatants and civilian populations present during and after the battles is much greater. However, these exposures have to be seen against the background of other exposures to potentially toxic agents associated with this war. Although exposure to DU may have played a role in the induction of any health effects demonstrated to have been induced, it may prove difficult to disentangle its effects in this multiple exposure situation and make clear attributions of specific health consequences to specific agents. Nevertheless, continued exposure to re-suspended DU/RU dusts could have posed and could continue to pose, a health hazard to the civilian population in the regions affected by the hostilities. As the soluble component is "weathered" away the risks will tend to converge towards those predicted on the basis of the ICRP lung model, taking into account the particle size distribution and any influence of the bystander effect.


ARCHER, V. E., 1981, Health concerns in uranium mining and milling. Journal of Occupational Medicine, 23, 502-505.

ARCHER, V. E., WAGONER, J. K. and LUNDIN, F. E., 1973, Lung cancer among uranium miners in the United States. Health Physics, 25, 351-371.

BAVERSTOCK, K. F. and THORNE, M. C., 1989, Radiological protection and the lymphatic system: the induction of leukaemia consequent upon the internal irradiation of the tracheobronchial lymph nodes and the gastrointestinal tract wall. International Journal of Radiation Biology, 55, 129-140.

BELYAKOV, O. V., MALCOLMSON, A. M., FOLKARD, M., PRISE, K. M. and MICHAEL, B. D., 2001, Direct evidence for a bystander effect of ionizing radiation in primary human fibroblasts. British Journal of Cancer, 84, 674-679.

BURKART, W. and LINDER, H., 1987, Hot particles in the environment: assessment of dose and health detriment. Sozial- und Praventivmedizin, 32, 310-315.

CARDIS, E. and RICHARDSON, D., 2000, Invited editorial: health effects of radiation exposure at uranium processing facilities. Journal of Radiological Protection, 20, 95-97.

CHECKOWAY, H., PEARCE, N., CRAWFORD-BROWN, D. J. and CRAGLE, D. L., 1988, Radiation doses and cause-specific mortality among workers at a nuclear materials fabrication plant. American Journal of Epidemiology, 127, 255-266.

CHPPM, 2000, Follow-up DoD Exposure Report; Depleted Uranium in the Gulf II, US Department of Defence. Available at:

COEN, N., MOTHERSILL, C., KADHIM, M. and WRIGHT, E. G., 2001, Heavy Metals of Relevance to Human Health Induced Genomic Instability. In Press.

DUPREE, E. A., CRAGLE, D. L., MCLAIN, R. W., CRAWFORD-BROWN, D. J. and TETA, M. J., 1987, Mortality among workers at a uranium processing facility, the Linde Air Products Company Ceramics Plant, 1943-1949. Scandinavian Journal of Work, Environment & Health, 13, 100-107.

IARC, 1990, Nickel and Nickel Compounds . Lyon, France, IARC.

ICRP, 1991, Recommendations of the International Commission on Radialogical Protection, Publication 60, Annals of the ICRP

ICRP, 1994, Human Respiratory Tract Model for Radiological Protection Publication 66, Annals of the ICRP. 24 (nos 1 - 3).

ICRP, 1995, Age-dependent Doses to Members of the Public from Intake of Radionuclides: Part 3 - Ingestion Dose Coefficients. Publication 69 Annals of the ICRP. 25(no 1).

KATHREN, R. L., MCINROY, J. F., MOORE, R. H. and DIETERT, S. E., 1989, Uranium in the tissues of an occupationally exposed individual. Health Physics, 57, 17-21.

KATHREN, R. L. and MOORE, R. H., 1986, Acute accidental inhalation of U: a 38-year follow-up. Health Physics, 51, 609-619.

KURTTIO, P., AUVINEN, A., SALONEN, L., SAHA, H., PEKKANEN, J., MÄKELÄINEN, I., VÄISÄNEN, S.B., PENTTILÄ, I.M., KOMULAINEN, H., in press, Renal effects of uranium in drinking water. Environmental Health Perspectives, in press.

LEGGETT, R. W., 1989, The behavior and chemical toxicity of U in the kidney: a reassessment. Health Physics, 57, 365-383.

LOOMIS, D. P. and WOLF, S. H., 1996, Mortality of workers at a nuclear materials production plant at Oak Ridge, Tennessee, 1947-1990. American Journal of Industrial Medicine, 29, 131-141.

MCDIARMID, M. A., 2001, Depleted uranium and public health (editorial). British Medical Journal, 322, 123-124.

MCGEOGHEGAN, D. and BINKS, K., 2000, The mortality and cancer morbidity experience of workers at the Springfields uranium production facility, 1946-95. Journal of Radiological Protection, 20, 111-137.

MILLER, A. C., BLAKLEY, W. F., LIVENGOOD, D., WHITTAKER, T., XU, J., EJNIK, J. W., HAMILTON, M. M., PARLETTE, E., John, T. S., GERSTENBERG, H. M. and HSU, H., 1998, Transformation of human osteoblast cells to the tumorigenic phenotype by depleted uranium-uranyl chloride. Environmental Health Perspectives, 106, 465-471.

MILLER, A. C., FUCIARELLI, A. F., JACKSON, W. E., EJNIK, E. J., EMOND, C., STROCKO, S., HOGAN, J., PAGE, N. and PELLMAR, T., 1998, Urinary and serum mutagenicity studies with rats implanted with depleted uranium or tantalum pellets. Mutagenesis, 13, 643-648.

MOTHERSILL, C. and SEYMOUR, C., 2001, Review: Radiation-induced Bystander Effects: Past History and Future Directions. Radiation Research, 155, 759-767.

RITZ, B., 1999, Cancer mortality among workers exposed to chemicals during uranium processing. Journal of Occupatioanl and Environmental Medicine, 41, 556-566.

RITZ, B., 1999, Radiation exposure and cancer mortality in uranium processing workers. Epidemiology, 10, 531-538.

RITZ, B., MORGENSTERN, H., CRAWFORD-BROWN, D. and YOUNG, B., 2000, The Effects of Internal Radiation Exposure on Cancer Mortality in Nuclear Workers at Rocketdyne/Atomics International. Environmental Health Perspectives, 108, 743-751.

SHEPPARD, S. C., and EVENDEN, W. G. 1988, Critical compliation and review of plant/soil concentration ratios for uranium, thorium and lead. J. Env Radioact. 8 255 - 285

RS, 2001, The Health Hazards of Depleted Uranium Munitions, Part I, The Royal Society, London, UK.

UNEP, 2001, Depleted Uranium in Kosovo, Post-Conflict Environmental Assessment, Switzerland.

VAN KAICK, G., MUTH, H., KAUL, A., WESCH, H., IMMICH, H., LIEBERMANN, D., LORENZ, D., LORENZ, W., LÜHRS, H., SCHEER, K. E., WAGNER, G. and WEGNENER, K., 1986, Report on the German Thorotrast Study.The Radiobiology of Radium and Thorotrast, Munich, Urban und Schwarzenberg.

WHO, 2001, Depleted Uranium, Sources, Exposure and Health Effects, World Health Organisation, Protection of the Human Environment, Geneva, Switzerland.