Canada’s Nuclear Regulator

Health Effects, Dosimetry and Radiological Protection of Tritium

Objectives

• Conduct an independent review of scientific literature to assess the adverse health risk to workers and the public from exposures to tritium.

• Assess Canadian and international dosimetry practices for tritium intakes.• Review the current approaches for limiting exposure to tritium.

To meet these objectives, the report provides:

• An overview of tritium’s physical, chemical and radiological properties • A detailed analysis of the adverse health effects of tritium radiation, including

reviews of laboratory and epidemiological studies• A review of experimental studies estimating the relative biological

effectiveness (RBE) of tritium radiation• A description of biokinetic models and dosimetry of tritium• A review of the approach taken by the International Commission on

Radiological Protection (ICRP) for protection from tritium and possible modification of the radiation weighting factor (wR)

Overall Study Conclusions

1. The lines of evidence, based on both epidemiological and laboratory studies, reveal that adverse health effects due to tritium exposure at the current exposure levels in Canada are highly unlikely.

2. The results of more than 50 experimental studies related to the determination of a single RBE value for tritium confirm that tritium beta radiation is about 1.4 times more biologically effective than radiation from 250 kVp x-rays and 2.2 times more biologically effective when compared to gamma rays.

3. Current dosimetry and biokinetic models for assessing dose are acceptable for radiation protection purposes.

4. The current regulatory framework, which uses the linear non-threshold risk model and the principle of ALARA (a practice that aims to keep radiation doses As Low As Reasonably Achievable) is satisfactory for controlling tritium exposures.

This study was peer reviewed by the following subject matter experts:

Dr. John Boice Professor of Medicine, Vanderbilt University School of MedicineScientific Director, International Epidemiology Institute

Dr. John HarrisonRadiation Biologist, Health Protection Agency, U.K.

Dr. David Kocher Senior Scientist, SENES Oak Ridge, Oak Ridge, Tennessee, U.S.A.

Dr. Barrie LambertRadiation Biologist, St Bartholomew’s Hospital Medical College, London, U.K. (retired)

Dr. Richard RichardsonInternal Dosimetrist, Atomic Energy of Canada Ltd., Chalk River, Ontario, Canada Dr. Richard Wakeford Visiting Professor in Epidemiology, Dalton Nuclear Institute, University of Manchester, U.K.

Canada’s Nuclear Regulator

Health Effects – Laboratory StudiesObjectives

The health effects of tritium have been studied extensively, perhaps more than any other radioisotope.

This study reviewed scientific literature on:

• Deterministic effects: changes in cells and tissues that are certain to occur after an acute dose of radiation (such as teratogenic (foetal) effects, reproductive effects and death)

• Stochastic effects: radiation-induced health effects (such as cancer and hereditary effects) where their probability of happening increases proportionally to the radiation dose received

• Organically bound tritium: tritium in molecules such as proteins, fats and starches

Main Findings

• Tritium behaves much the same as other ionizing radiations in inducing deterministic effects.

• Laboratory studies using animals have demonstrated that tritium exposures, as with other sources of radiation exposure, can interfere with the development of an embryo or foetus, promote cancer, and induce genetic and reproductive effects and cell death.

• The actual quantity of tritium required to induce these effects is very large; in the order of gigabecquerels (GBq; billions of tritium atoms decaying and emitting a beta particle per second) per gram of body weight. The dose resulting from these intakes is more than 500 millisieverts (mSv).

• Organically bound tritium is more effective at causing adverse health effects than tritiated water (HTO), primarily due to the length of time spent in the body (biological half-life). There are exceptions, but these are not a risk in environmental settings.

Table 1 summarizes the studies reviewed in this report, although care needs to be taken in comparing the endpoints and different dose regimes (injection or ingestion), especially with respect to the mouse oocyte. Generally, doses above 0.5 GBq and radiation doses of 0.5 Gy are needed to induce an adverse health effect. The exception to this is the mouse reproductive system, whose oocytes are about 100 times more radiosensitive than that of humans.

Table 1. Overview of studies showing lowest concentration or administered dose of tritum to produce an adverse effectStudy

LethalityBrues et al (1952)

Yokoro et al (1986)

Yokoro et al (1986)

Furchner (1957)

Yamamoto et al (1990)

Zamenhof & Van Marthans (1979)

Laskey et al (1973)

Bursian et al (1975)

Jones et al (1980)

Satow et al (1989a,b)Lambert (1969)

Lambert (1969)

CarcinogenicityYamamoto et al (1995)Yamamoto et al (1995)Yamamoto et al (1998)Gragtmans et al (1984)Johnson et al (1995)Seyama et al (1991)Mushkacheva et al (1992)

Means of Exposure

Single injection of HTO

Single injection of HTOSingle injection of HTOSingle injection of HTOIngestion of HTO

Ingestion of HTO

Ingestion of HTO

Ingestion of HTO

Ingestion of HTO

Injection of HTOInjection of 3HTdRInjection of HTO

Ingestion of HTO

Ingestion of HTO

Ingestion of HTO

Injection of HTO

Injection of HTOInjection of HTOIngestion of HTO

Test Animals

Mice

Mice

Mice

Mice

Mice

Mice

Mice

Mice

Squirrel monkeys

MiceMice

Mice

Mice

Mice

Mice

Rats

MiceMiceMice

Parameter

LD50/30

Lowest dose

LD50/30

LD50/30

Lowest dose

No observed effects in mothers, temporary effects in offspring Lowest dose Reduced brain weightreduced brain weight and neural hormones Oocyte reduction, no other noted effects

Oocyte reductionReduction of spermatogoniaReduction of spermatogonia

Thymic lymphoma

Solid tumours

Thymic lymphoma

Early onset of mammary tumour Myeloid leukaemiaTumoursDNA damage

Lowest Concentration or Administered Dose to Show Effect

0.37 GBq/g body weight (BW) (~1.5 GBq total dose)0.37 GBq

0.56–0.93 GBq

0.035 GBq/g(~1.3 GBq total dose)1.37 GBq

111 MBq/L

3.7 MBq/L

370 MBq/L

111 MBq/L

0.17 MBq/g BW185 kBq/g BW

2.22 MBq/g BW

1.85 GBq/L

0.925 GBq/L

0.463 GBq/L

0.45 MBq/g BW

90 MBq140 MBq37 kBq/g/d

Dose (When Available)

N/A

~5 Gy

~8–13 Gy

N/A

11.1 Gy

N/A

0.00003 Gy/d

0.66 Gy

N/A

0.035 Gy0.084 Gy (to cell nucleus)0.049 Gy (to cell nucleus)

0.048 Gy/d

0.024 Gy/d

0.012 Gy/d

0.46 Gy

0.85 Gy1.97 Gy1.2 Gy

Teratogenic effects

Reproductive effects

Health Effects, Dosimetry and Radiological Protection of Tritium (continued)

Canada’s Nuclear Regulator

Epidemiology is the study of the factors affecting the health and illness of populations and how disease is distributed. It serves as the foundation for public health and preventive medicine.

Studies based on good-quality radiation exposure data provide the best source of evidence for estimating human health risks from radiation.

Main Findings

• Many epidemiological studies involving radiation workers, their offspring and members of the public living near nuclear facilities, and several major authoritative reviews of the scientific literature have been reviewed.

• To date, the existing information does not contain enough detail to specifically estimate the health risks of tritium exposure.

• Tritium exposures represent a very small fraction of total radiation exposures from all sources of radiation.

• Based on the lack of excess health risks found from total radiation exposures (which are low) among the populations studied, any health risks to workers and the general population from tritium are expected to be negligible and not distinguishable from the risks of similar health outcomes in the general population.

• To date, no human health studies have demonstrated any tritium-induced cancer or other health effect.

Table 2 is a summary from the study by Zablotska et al (2004), which involved a mortality follow-up of 45,468 Canadian nuclear power industry workers from 1957 to 1994. Overall, there was no indication of any unexpected pattern in the cohort mortality compared with that of the general population.

Table 2. Summary of cancer mortality excess relative risk results from Zablotska et al (2004)

Background rate stratified by sex, age, calendar year, SES and duration of monitoringa Doses lagged by 2 yearsb Doses lagged by 10 yearsc Negative trend

Studies of populations living near nuclear facilities have major limitations (such as unknown exposure to other carcinogens and unmeasured exposures to tritium) and are unlikely to add to our understanding of the health effects of tritium. An international collaborative study combining tritium worker data from multiple countries would be required to provide the necessary study power to assess tritium risk directly.

Disease

All non-CLL leukaemiaa

All leukaemiaa

All solid cancersb

Rectal cancerb

Lung cancerb

Non-Hodgkin’s lymphomab

Multiple myelomab

Pancreatic cancerb

Brain and other central

nervous system cancerb

Number of

Observed Deaths

18

22

474

16

183

19

10

22

25

Excess Relative

Risk per Sv

52.5

18.9

2.80

34.1

4.34

-2.03

-2.04c

-2.03

-2.04

95% Confidence

Intervals

0.205–291

<-2.08–138

-0.038–7.13

1.41–165

-0.193–12.7

<-2.08–38.0

<-2.08–18.3

<-2.08–34.4

<-2.08–8.99

2-Sided

p-Values

p=0.048

p=0.25

p=0.054

p=0.029

p=0.066

p=0.74

p=0.41

p=0.59

p=0.34

Health Effects, Dosimetry and Radiological Protection of Tritium (continued)Health Effects – Epidemiology

Canada’s Nuclear Regulator

Health Effects, Dosimetry and Radiological Protection of Tritium (continued)Relative Biological Effectiveness

• Relative biological effectiveness (RBE) is a relative measure of the effectiveness of different radiation types in inducing a specified health effect. It is defined as:

absorbed dose from a given radiation absorbed dose from the reference radiation

• The RBE contributes to the basis for a universal measurement of dose, the sievert (Sv), for radiation protection purposes. It is also useful for retrospective studies where a more exact dose measurement is needed due to a high exposure requiring medical treatment or for epidemiological studies requiring the best estimate of each individual dose.

• The choice of a single value for the RBE of tritium for radiation protection purposes is difficult. There have been more than 50 different estimates of a tritium RBE due to considerable variation and uncertainty in the radiobiological data.

• The choice of a reference radiation makes up much of this variability, since the RBE for x-rays and gamma rays—the two types of radiations typically used as the reference radiation—differ by nearly a factor of two. The review of studies to determine a single value for the RBE of tritium radiation indicates that:

- If x-rays are the reference radiation, an RBE value of about 1.4 for tritium would be appropriate.

- If gamma radiation is chosen as the reference, an RBE value closer to 2.2 would be appropriate.

- Gamma radiation appears to be the preferred reference radiation (ICRP, 2003) principally because the atomic bomb survivors were primarily exposed to gamma radiation.

Table 3 provides a statistical summary of the RBE determinations under many different experimental variables.

Table 3. Statistical Summary of Studies Determining the RBE of Tritium Radiation under

Different Experimental Conditions (95% confidence interval provided in brackets)

* Indicates values were calculated for this report.

Main Findings

• There is no significant difference between in vivo and in vitro studies.• There are significant differences in RBE values for the reference radiation and if the reference radiation was delivered chronically or acutely. • The studies with a chronic gamma reference radiation have a mean RBE of about 30% higher than those that used chronic X-ray as the reference radiation.

All in vivo

All in vitro

In vivo acute X-ray ref.

In vivo chronic X-ray ref.

In vivo chronic gamma ref.

In vitro gamma ref.

In vitro acute X-ray ref.

All chronic X-ray ref.

All acute X-ray ref.

All chronic gamma ref.

All acute gamma ref.

All chronic

All acute

All X-ray ref.

All gamma ray ref.

RBE all studies combined

Cancer studies with

chronic gamma

Cancer studies with

chronic X-ray

All except studies of

survival and inactivation,

with chronic gamma ray

All except studies of

survival and inactivation,

with chronic X-ray

Mean (95%

Confidence

Interval)

1.84 (1.49–2.18)

1.85 (1.61–2.08)

0.81 (0.46–1.17)

1.34 (0.94–1.74)

2.36 (2.03–2.68)

1.91 (1.56–2.27)

1.80 (1.39–2.20)

1.34 (0.94–1.73)

1.28 (0.93–1.63)

2.15 (1.89–2.41)

Insufficient data

2.00 (1.75–2.25)

1.28 (0.53–1.64)

1.44 (1.15–1.73)

2.15 (1.90–2.41)

1.88 (1.66–2.09)

Mean with Little

and Lambert

Recalculations

1.75 (1.40–2.11)

2.14 (1.45–2.84)

Insufficient data

Insufficient data

2.21 (1.80–2.72)

2.11 (1.50–2.72)

2.34 (0.94–3.74)

Insufficient data

1.56 (0.70–2.42)

2.16 (1.80–2.52)

Insufficient data

2.01 (1.69–2.33)

1.56 (0.70–2.42)

1.69 (0.97–2.42)

2.16 (1.81–2.51)

2.01 (1.66–2.37)

Straume and

Carsten

(1993)

1.67* (1.31–2.03)*

1.69* (1.26–2.12)*

2.36* (1.96–2.76)*

1.95* (1.14–2.77)*

1.8 (1.44–2.16)*

2.3 (1.94–2.67)*

2.11* (1.82–2.40)*

RBE=Little and

Lambert

(2007)

0.56 (0.31–0.96)

1.17 (0.96–.39)

1.63 (1.49–1.77)

1.98 (1.85–2.12)

1.45 (1.32–1.58)

1.17 (0.96–1.39)

2.19 (2.04–2.33)

2.49 (2.00–2.98)

1.19 (0.88–1.49)

2.19 (2.04–2.33)

1.17 (0.96–1.39)

Canada’s Nuclear Regulator

• Dosimetry is a scientific subspecialty in radiation protection and medical physics that focuses on calculating the dose resulting from exposure to ionizing radiation.

• Dose cannot be measured directly with instruments; instead, it must be calculated from measurements of the amount of radiation to which a person has been exposed.

• Dose is proportional to the amount of radiation energy absorbed by the body per kilogram of body weight.

• Doses from tritium are usually determined from bioassay samples (such as urine) of occupationally exposed individuals or through environmental monitoring (e.g., tritium in air).

ICRP Models

To calculate the dose resulting from tritium exposure, it is first necessary to know how much tritium is retained in the various organs and tissues of the body over time after exposure to tritium.

The International Commission on Radiological Protection (ICRP) is the international organization that makes recommendations for radiation protection based upon current assessments of health effects caused by radiation exposure. It has recommended two models to calculate the dose from tritiated compounds:

1. The ICRP HTO (tritiated water) model (see Figure 1) is used to calculate the dose resulting from intakes of tritiated water or other tritiated compounds that partially convert to HTO after being taken into the body.

2. The ICRP OBT (organically bound tritium) model (see Figure 2) is used to

calculate the dose resulting from intakes of various tritiated organic compounds (such as nutrients). The ICRP OBT model applies to the inhalation and ingestion of organically bound tritiated compounds. This model predicts that for equal amounts of tritium ingested, organically bound tritium yields about twice the dose compared to that from tritiated water.

Studies have shown that these ICRP models are reasonably consistent with experimental results. The following are a graphic description of the current ICRP models.

Figure 1. ICRP model for the biokinetics of tritiated water

Figure 2. ICRP model for the biokinetics of organically bound tritium

It is important to note that there are ongoing studies looking at improving the accuracy of these models, based upon differences in retention time (or biological half-life).

Because tritiated water largely has the same chemical and physical properties as normal water, it will be distributed evenly throughout the body as would any water molecule. Its retention in the body is also the same as any other water molecule.

However, in most OBT compounds, tritium that is bound to oxygen, nitrogen, phosphorus or sulphur will exchange freely with hydrogen in water. Therefore, it has the same metabolism and distribution in the body as HTO and is called the exchangeable bound tritium fraction. Tritium bound to carbon will not exchange with hydrogen in water and is known as non-exchangeable bound tritium. Biokinetic models under consideration attempt to better estimate the differences in retention time for the exchangeable and non-exchangeable OBT compounds.

Blood

Excreta

HTO OBT

Blood

HTO OBT

GI Tract

97%

Urine (47%) Feces (3%) Other (50%)

3%

50% 50%

T1/2=10 days T1/2= 40 days

Excreta

Urine (47%) Feces (3%) Other (50%)

T1/2=10 days T1/2= 40 days

Health Effects, Dosimetry and Radiological Protection of Tritium (continued)Dosimetry

Canada’s Nuclear Regulator

Nationally and internationally, the approach taken to protect radiation workers and members of the public from radiation is to adopt the principles and recommendations of the International Commission on Radiological Protection (ICRP).

The ICRP has designated the sievert (Sv) as the unit to give a measure of dose for all ionizing radiations. It does this by applying weighting factors (wR) for the different types of radiation (alpha, beta and gamma) and for the radiation sensitivities of different organs and tissues.

• The sievert is strictly a unit for radiation protection purposes.• It provides a single unit for dose from all ionizing radiations for optimization

and to compare against the dose limit.• Due to simplifications, the wR loosely reflects the biological effectiveness of

the type of radiation and is therefore only an approximate indicator of risk. • All electrons (beta radiation) and photons have a weighting factor of 1.• The wR is not based upon the source of the radiation (for example, x-ray

machines or specific radioisotopes).• Doses are gender neutral; equivalent and effective doses are calculated for a

‘representative person’ based on a population of males and females, ethnicity and age.

• The sievert should not be used to assess doses when individual risk assessments are required, such as cases where large intakes are suspected or in epidemiological studies. In those cases, tissue-weighted absorbed doses with appropriate relative biological effectiveness (RBE) values should be used. As well, characteristics specific to the individual should be taken into account.

The ICRP does not believe that practical radiation protection would be improved by having radioisotope-specific radiation weighting factors or separate dose estimates for males and females. But for its recommendation to be effective, doses must be kept As Low As Reasonably Achievable (ALARA).

Using a different wR for tritium to reflect the RBE value would best reflect the radiation risk from tritium. However, the following considerations must be taken into account with this approach:

• The apparent improvement in correlation with risk may be misleading given the existence of many other uncertainties due to other significant variables (sex, age, weight, metabolic rates, genetic susceptibilities, etc.). Occupational and public doses are already low and there is a continuous effort to reduce these further.

• It would be inconsistent with the ICRP system of radiological protection which is currently the international standard approach. Specifically:

a) There are no other isotope-specific wR values.b) It would be difficult to compare radiation protection practices nationally

and internationally.

Health Effects, Dosimetry and Radiological Protection of Tritium (continued)Options for Assessing and Controlling Risks Associated with Tritium Exposures