Radiation Dose Units: How We Got Here,Do They Serve the Purposes They are Intended to?

Abstract

The current radiation dose estimates used in medical imaging, radiation oncology or environmental assessments are not entirely accurate from a fundamental physics perspective, let alone for biological consequences. The “one cloth fits all” approach of radiation-matter interactions cannot assess the effects of interactions of the same species of radiation of different energies on the same isotope of an element. Preliminary steps to set the radiation dosimetry in the right direction are suggested.

Discussion

Societies experienced human fatalities and sicknesses due to radioactive substances well before nuclear radiations were known. The occupational hazards of working in mines was felt as early as in the 15^th Century in Schneeberg, Germany and Joachimsthal (Jachymov) in Czechslovakia (present-day’s Czech Republic). ¹ Many miners in both areas had shortened life due to a Bergkrankheit (mountain sickness). It was perhaps the first known radiation sickness. However, it was not until 1930s, radium and radon inhalations were finally acknowledged as the causes for lung cancers. The old adage, ‘What you don’t know can’t hurt you’ ² does not seem to apply to radiation. Ignorance is no bliss when we deal with radiation.

While the hazards of X-rays were recognized quite early on due to skin burns, the discovery of radium and uranium did not result in quick realization of radiation hazards. It was only in the 3^rd decade of the 20^th Century, that protection against the hazards of X-rays and radium were articulated and a tolerance dose was proposed in 1925 by Arthur Mutscheller. ³

Well-known was the “radium girls” saga of 1920s, when several young girls employed at radium dial painting jobs suffered cancers and died. Also, it was the time, when ‘Radithor’ solution was promoted as “a cure for the living Dead”. Radithor is a mixture of ²²⁶Ra and ²²⁸ Ra (aka mesothorium). Three to four daily doses of radithor containing 1 mCi each of the two isotopes was recommended. Mr. Eben Byers, a celebrity industrialist, succumbed to death as he consumed four daily doses of medical nostrum. This news was a catalyst for the engagement of Robley Evans ⁴ to assess radiation hazards, who later established an active research group at MIT. The U.S. X-ray and Radiation Protection Committee, with Evans as a member, noted that the radioactivity levels of 1 nCi/L are carcinogenic and a safety factor of 100 led them to recommend 10 pCi/L as the acceptable dose limit ¹ . There are a few technicalities to note here: (a) This recommendation does not take the energies and number of emanations in each decay, nor does it consider the sequential daughter products of the parent nuclei and their emissions. (b) At best, it is a criterion that only serves the dosage effects of radium. However, it is recognized that the ingestion of radioactive sources can have life-long effects as the emitters, once ingested, may remain in human bodies for indeterminate times which vary with the half-lives and chemical properties and the sites they accumulate in.

Barium and calcium, which are known to settle on bone, belong to the same chemical group as ²²⁶Ra which has a physical half-life of 1600 years, more than an order of magnitude longer than a human life span. Based on these numbers, researchers extrapolated backwards a recommended safe dose of 0.1 mCi. This is the amount of isotope that a person may ingest and still be safe and was less than 1/5^th of what was deemed to be the lowest hazard limit, proving to be a safe limit without adverse consequences. ²

Of note, the ‘radium girls’ effects indicated that 1 nCi/L radium was carcinogenic. If one estimates the total amount of alpha particle energies (including the energies of all daughter products of the chain) to be approximately 30 MeV/Bq, the energy deposit of 1 nCi/L (37 Bq/L) corresponds to about 10⁻¹⁰ Gy/s. The present-day brachytherapy treatments are carried out with a few gigabecquerel (∼Ci) activities ⁶ which subject patients to doses of a few grays, dose rates much higher than the annual dose of radium girls. If one considers radioactive levels or energy deposits as the sole measures of radiation hazards, present-day medical techniques of radiation therapies and diagnostics would be impossible.

Until the landmark work of Lorenz, ⁷ the role of the daughter products went unnoticed. Radon itself contributed about 2% of the radiation dose, while the daughters (²¹⁴Bi-RaC and ²¹⁴Po- RaC′) accounted for the rest. Radiations from these decays of short-lived isotopes will persist as long as ²²⁶Ra is present. Since ²²²Rn is noble gas, its progenies can migrate to other sites. In their evaluations, Evans and his contemporaries, despite their meticulous endeavours, missed to assess the contributions of subsequent emissions for their radiation effects. Also, the tolerance dose, the maximum amount of radiation an individual maybe exposed to, was decided very subjectively by a committee with what they could be comfortable to have their wife and daughters be exposed to (Ref. 4). In 1977, the International Commission of Radiation Protection (IRPC) introduced a risk-based criterion that radiation workers should not be at more risk than an average worker in a ‘safe’ industry would be subject to (Ref. 5). To the best of our knowledge, the present-day regulations are based on the same principle and that the public should be exposed to less than that of a radiation worker.

During the intervening decades, a thesis of “all radiations are harmful” was dominant with an assertion of linear no-threshold (LNT) for radiation damage and doses as low as reasonably achievable (ALARA) guide our dealings with nuclear radiations. This is despite several professionals countering these theses with some observational data and reasonings that a finite non-zero low doses can have hormetic effect. Among others, Dr T.D. Luckey proclaimed that dose is everything and that “large doses are harmful; small doses are stimulatory”. ⁷ Even before the hormesis effect was suggested, there were studies to refute the LNT model while they did not claim beneficial aspects of radiation. ⁸ Nearly 50 years later, in 2024, the debates of LNT vs non-LNT and Hormesis vs no-Hormesis continue. The data extrapolations, interpolation etc., were also characterized as “creative statistics and statistical fantasy” (ref. 6, p. 226).

We believe that all professionals endeavour to present the best possible data and their interpretation with a goal to save humanity and ecology. To our thinking, the ongoing saga of bitter debates on radiation hormesis, LNT and other non-LNT models is due to a problem which is much deeper and fundamental than that of shortage of samples, measurements and statistics. We think it is unfortunate that energy deposited by the radiations has been deemed as the basic measure of radiation effects. To some extent, one can rationalize this criterion in times before the advent of particle accelerators, when the effects were limited to radiation interaction with atoms or molecules inducing ionization or electron displacements. These interactions vary nearly monotonously with changes of atomic number of the interacting medium across the periodic table except for chemical toxicity. These effects could be reasonably modeled by the amount of energy deposits.

In modern times, when electron, proton and heavy ion beams of multi million electron volts (MeV) are used in radiation therapy and particle accelerators of giga and tera electron Volts are employed in research, one cannot ignore nuclear transmutations induced in material media. These transformations are stochastic and quite often cause permanent changes in the medium (e.g. organs). Nuclear processes are sensitive not only to the elemental content of the material medium but depend on the isotope compositions. If the species and energies of particle beams or elemental isotope content of the medium changes, nuclear transformations may vary drastically and in an unpredictable manner. In addition, the nuclear transformations are irreversible changes in the atomic/isotopic composition of a medium and may be irreparable. A simple example will be interaction of neutrons with uranium. Thermal neutrons of very low energies easily cause nuclear fission of uranium-235 and are useful as a source of energy. However, much more energetic neutrons are required to cause fission of uranium-238 and it is yet to be exploited as energy resource. Two observations are in order: (a) Energy deposit of a species of radiation is not an absolute measure of dose imparted, (b) a single parameter of radio biological effect (RBE) cannot capture the medium and incident beam dependent effects.

Below is a simple illustration of photon interactions in water, a model medium for the human body.

In recent years, photon energies in excess of 20 MeV are being employed for radiation therapy. Figure 1 shows cross section data of interactions of photons of up to 25 MeV^9,10 for attenuation in water and photo nuclear reactions in ¹⁶O. For photon energies below 7 MeV, no nuclear transmutation of ¹⁶O is possible and all the interactions and energy deposits are due to monotonous atomic processes (blue curve in the figure). Above 7 MeV, transmutation of oxygen to carbon by emission of alpha particle is energetically possible, but it is negligibly small for physics reasons. The real onset of transmutation occurs at above 15 MeV when emissions of neutrons and protons become energetically possible. The nuclear transmutation increases by about a factor of 5 and exhibits a broad maximum at about 20 MeV energies (red circles in the figure). It should be noted that energy deposits due to residuals, protons, neutrons etc. are smaller compared to atomic/molecular interactions. However, the transmutations are permanent, irreversible changes in the water molecule. While this is one example, it should be remembered that the human body is a composite of several elements each of which have their unique energy-dependent nuclear responses to each species of radiation they encounter. Neither the gray nor sievert as energy deposit measures and the corresponding biological equivalent corrections can incorporate these manifestations.

Figure 1.

Water-photon interactions as a function of photon energy. Blue data is atomic/molecular interaction data from NIST’s XCOM database ⁹ and takes into account ionization, scattering, and pair production. Red circles are nuclear data. ¹⁰ See text for details.

Recently, Masuda et al. ¹¹ reported their measurement of proton interactions in water for up to 70 MeV energies detecting several radioactive isotopes (¹³N, ¹⁵O, …) and simulated the corresponding energy distributions of protons in medium. It is desirable to extend their work to assess changes in the elemental composition of medium. Present-day proton therapy uses beams of up to 250 MeV energies, at which several other reaction channels are open with more energetic reaction products, spreading over larger volumes in the medium. In addition to nuclear disintegrations, meson production becomes possible. The commonly estimated KERMA can only account for the energy deposits due to secondaries. For example, the secondaries in oxygen-proton interactions may be carbon, boron and beryllium among others. Excessive boron or beryllium is toxic to humans. Beryllium belongs to the same group as calcium may attach to bones, just as radium does. Also, the living organisms consist of other elements such as carbon, calcium, iron, to name just a few. The above discussion about photon-water interactions is meant for illustrative purposes only. It is not intended to be an exhaustive description of all dose effects, but to highlight that energy deposit estimates fail to account for radiation dose effects even at physics level, let alone the biological consequences.

To summarize, we reason that present-day radiation dosimetry based on absorbed energy, expressed as the gray, weighted with a quality factor, to constitute the sievert, as the dose estimate cannot fully capture the products of complex nuclear reactions and their consequences. While we do not have a precise solution, we propose the following approach as a step forward:

A) Consider water medium as is usually done.

B) Evaluate, for each species of radiation (neutrons, gammas, protons, electrons, etc….) the role of energy dependent processes and consequences to the physico-chemical composition of the medium.

C) Present these results to the stakeholders as beam energy specific neutron-water dose, or photon-water dose etc.

D) As for the future steps, incorporate other elements Ca, K, Fe etc., present in human organs for dose assessments.

In recent years, real time monitoring of mixed radiation fields has become possible and is evolving. ¹² It is a simple next step to convert this data into the partial doses due to diverse radiations and present the radiological effect on a human being of known physiological composition in real time.

This distinction would be very important for several reasons. As an example, in case of a nuclear event, the radiation dose may be a composite of neutrons, photons and fission fragments. A few minutes or hours later, neutrons would become non-existent either by decay or due to interactions with the surrounding media. The remainder is mainly photons and radioactive emissions of fission fragments, if any. In case of radiation procedures on humans or other living organisms, there are again neutrons and other radiations during irradiation. Later, it is emissions from residual radionuclides. The energies of these secondary radiations are less than about 3 MeV, below the threshold of inducing nuclear processes and transmutations. Thus, the radiation effects of later times are different from those of early times of nuclear event. Also, in a medical procedure, one can assess the of the prompt radiations due to the primary beam interactions with atoms, molecules and atomic nuclei. Biological effects of post-treatment are due to secondary and tertiary emissions, which may be quite different from the primary beam’s energy loss mechanisms. These specifications may be improved and simplified as we gain better knowledge.

Conclusion

Despite the long history of our awareness of radiation hazards, accurate quantification of radiation effects remains elusive not just for biological consequences but also for physico-chemical transformations. With changing times and evolving technologies, we continue to develop sources of diverse types of radiations of wide range of energies for multiple uses. It is pointed out that the complex radiation-material interactions cannot be quantified simply in terms of energy deposits. Radiation dose units based on science of the early 20^th century for low energy radiations of pre-accelerator times are inadequate to quantify the radiation phenomena involving much higher energies. This is particularly true for interactions of radiation with atomic nuclei. The sievert, a multiplication factor of gray (energy deposit), falls short of the challenge of being a unit to quantify radiation effects.