Comparing absorbed doses and radiation risk of the α-emitting bone-seekers [223Ra]RaCl2 and [224Ra]RaCl2

[223Ra]RaCl2 and [224Ra]RaCl2 are bone seekers, emitting high LET, and short range (< 100 μm) alpha-particles. Both radionuclides show similar decay properties; the total alpha energies are comparable (223Ra: ≈28 MeV, 224Ra: ≈26 MeV). [224Ra]RaCl2 has been used from the mid-1940s until 1990 for treating different bone and joint diseases with activities of up to approximately 50 MBq [224Ra]RaCl2. In 2013 [223Ra]RaCl2 obtained marketing authorization by the FDA and by the European Union for the treatment of metastatic prostate cancer with an activity to administer of 0.055 MBq per kg body weight for six cycles. For intravenous injections in humans a model calculation using the biokinetic model of ICRP67 shows a ratio of organ absorbed dose coefficients (224Ra:223Ra) between 0.37 (liver) and 0.97 except for the kidneys (2.27) and blood (1.57). For the red marrow as primary organ-at-risk, the ratio is 0.57. The differences are mainly caused be the differing half-lives of the decay products of both radium isotopes. Both radionuclides show comparable DNA damage patterns in peripheral blood mononuclear cells after internal ex-vivo irradiation. Data on the long-term radiation-associated side effects are only available for treatment with [224Ra]RaCl2. Two epidemiological studies followed two patient groups treated with [224Ra]RaCl2 for more than 25 years. One of them was the “Spiess study”, a cohort of 899 juvenile patients who received several injections of [224Ra]RaCl2 with a mean specific activity of 0.66 MBq/kg. Another patient group of ankylosing spondylitis patients was treated with 10 repeated intravenous injections of [224Ra]RaCl2, 1 MBq each, 1 week apart. In total 1,471 of these patients were followed-up in the “Wick study”. In both studies, an increased cancer mortality by leukemia and solid cancers was observed. Similar considerations on long-term effects likely apply to [223Ra]RaCl2 as well since the biokinetics are similar and the absorbed doses in the same range. However, this increased risk will most likely not be observed due to the much shorter life expectancy of prostate cancer patients treated with [223Ra]RaCl2.


Introduction
[ 223 Ra]RaCl 2 targets bone metastases with high LET and short range (<100 µm) alpha-particles. In 2013, Parker et al. published the results of the phase III, double-blind, randomized, international ALSYMPCA study which compared [ 223 Ra]RaCl 2 plus best standard of care (BSC) vs. placebo plus BSC in castration resistant prostate cancer (CRPC) patients with bone metastases (1). The authors concluded that the ALSYMPCA study demonstrated significantly improved overall survival and low toxicity, suggesting that [ 223 Ra]RaCl 2 may provide a new standard of care for patients with CRPC and bone metastases. The results of the ALSYMPCA trial were used to obtain marketing authorization for [ 223 Ra]RaCl 2 ("XOFIGO" R ) in Europe and North America in 2013.
[ 224 Ra]RaCl 2 has been used from the mid-1940s until 1990 for treating different bone and joint diseases, mainly in Germany (2, 3). After World War II, [ 224 Ra]RaCl 2 was primarily used for the treatment of children and juveniles suffering from bone tuberculosis, and even for the therapy of Ankylosing Spondylitis (AS) patients. The activities of [ 224 Ra]RaCl 2 administered at that time were high (approximately 0.66 MBq/kg body weight, corresponding to an activity of 50 MBq), with treatment durations ranging from 1 month to 45 months (median: 4 months). In the "Spiess study" 899 patients who received multiple injections of [ 224 Ra]RaCl 2 mainly between 1945 and 1955 for the treatment of tuberculosis, AS and some other diseases had been followed (3).
In a second group of patients who were treated with repeated intravenous injections of [ 224 Ra]RaCl 2 (excluding radiation therapy with X-rays) between 1948 and 1975 an epidemiological study on 1,471 ankylosing spondylitis patients was performed ("Wick study"). The activity was administered as 10 intravenous (IV) injections, 1 MBq each, one a week apart (mean: 0.17 MBq/kg, 10 MBq total). These patients have been followed together with a control group of 1,324 AS patients treated neither with radioactive drugs nor with X-rays (2).

Radioactive decay and exposure
Decay chains Ra 223 Ra is an alpha emitter (half-life = 11.43 d), which decays through a cascade of short-lived alpha-and beta-emitting progeny with the emission of about 20 MeV of energy per starting atom and the first two daughters and about 28 MeV through complete decay of the progeny to stable lead ( Figure 1). A listing of the decay chain, branching ratios, half-lives, energies emitted by alpha-, beta, and gamma-transitions is provided e.g., by Schumann et al. (6). The data for the energy per transition in this publication was taken from the MIRD tables by Eckerman and Endo (7). 224 Ra is also an alpha emitter (half-life = 3.63 d) decaying through a cascade of short-lived alpha-and beta-emitting progeny with the emission of about 19 MeV of energy per starting atom and the first two daughters and about 26 MeV through complete decay of the progeny to stable lead ( Figure 2). More details on the decay chain and the energies emitted are provided by Schumann et al. (6) and were also taken from the Eckerman and Endo tables (7).
Several clinical studies measured the disappearance of [ 223 Ra]RaCl 2 from the blood and the excretion pathways (11)(12)(13)(14). All studies showed a rapid blood clearance; the major excretion pathway, however, is fecal excretion.

FIGURE
Decay chain of Ra. Decay products with branching ratios < % are omitted. The decay data were taken from http://www.nucleide.org/ Laraweb/index.php. To compare the dosimetry data for both radionuclides the absorbed dose coefficients were taken from the tables provided by Lassmann et al. (8,10). The data for blood were taken from Schumann et al. (14) and Stephan et al. (18).

Comparison of absorbed doses to organs or tissues
In Table 1 For most organs or tissues all decay products contribute almost equally to the absorbed doses in these organs (14,16). Experimental data on these effects, however, are sparse and are taken from animal experiments (19). For the red marrow as primary organ-at-risk, the ratio of the absorbed dose coefficients is 0.57. The largest dissimilarities of the absorbed dose coefficient ratios are observed for the kidneys (2.27), blood (1.67), and liver (0.37). The higher values for the kidneys and blood could be attributed to the accumulation of lead and its progeny due to the longer half-life of 212 Pb compared to 211 Pb.
A comparison of the absorbed dose ratios assessed for the two treatment scenarios shows that the absorbed doses are always lower for [ 224  This comparison does not include absorbed doses of metastases which take up radium as the underlying ICRP models do not consider this case as they were designed for radiation protection purposes. Therefore, the absorbed doses to organs/tissue could be much lower if a considerable amount of . /fmed. . the injected activity is taken up by tumors as only a fraction of the remaining activity will be available and taken up by other organs or tissues.

External exposure
To further elucidate potential differences between 224 Ra and 223 Ra and their respected progenies regarding exposure of staff or persons staying close to patients, the dose rate constants for the ambient dose H * were compared. For the comparison, the newest published values were used for 224 Ra and 223 Ra and the respective progenies (20). The

Long-term radiation-related e ects Patient cohorts studying long-term radiation-related e ects of [ Ra]RaCl
There are two patient cohorts that were followed for longterm radiation-related effects after the use of [ 224 1948 and 1975 (2). These patients have been followed in the "Wick study" together with a control group of 1,324 AS patients treated neither with radioactive drugs nor with X-rays. The mean follow-up time was 26.3 years in the exposed and 24.6 years in the control group.

Radiation-induced side-e ects of [ Ra]RaCl and [ Ra]RaCl
In the study cohort of the "Spiess study", Nekolla et al. For [ 224 Ra]RaCl 2 the most striking observation of the "Wick study" (2) were the 21 cases of leukemia in the exposed group .
(vs. 6.8 cases expected, P < 0.001) compared to 12 cases of leukemia in the control group (vs. 7.5 cases expected). This increase in total leukemias was significant in direct comparison between the exposed and control groups too (P < 0.05). Wick et al. found, besides an increased standardized incidence ratio (= ratio of the number of observed cases vs. the number of expected cases) of leukemias, a significant increase for kidney and thyroid cancer (2). For [ 223 Ra]RaCl 2 only mild side and mostly transient effects were observed (1). [ 223 Ra]RaCl 2 was well tolerated by patients with skeletal metastases. Mild to moderate and transient hematological toxicity was observed at potentially therapeutic doses. Platelets were less affected than neutrophils and white blood cells; toxicity grade I was seen in 5 of the 31 patients (1). Furthermore, only two cases of leukemia have been reported until today (24).

Discussion
A major drawback for image-based dosimetry of [ 223 Ra]RaCl 2 is the inherent difficulty to quantify posttherapeutic gamma camera images, although photon emissions suitable for gamma camera imaging are available at ∼82 keV, ∼154 keV, and ∼270 keV. Due to the low photon abundance, the low activities administered to the patients, and the high contribution of down-scatter of higher energy photons leading to severe septal penetration causes large image quantification uncertainties as reported by Hindorf et al. (25). Pacilio et al. (26) and Yoshida et al. (13) provided quantitative results by planar imaging, however, the accuracy of the respective quantification process for in-vivo imaging is even more limited due to activity overlay in this type of image. For 224 Ra, data on imaging, though theoretically possible with the 239-241 keV gamma rays for 224  A major concern for the application of radium isotopes to patients could be diffusion of the first daughter products 219 Rn (half-life: 4 s) or 220 Rn (half-life: 56 s). This could lead either to an increased diffusion of radon away from the binding site leading to unwanted irradiation of other organs or tissues or to increased emanation of radon, therefore reducing the energy deposited in the tumor/lesion. Lloyd et al. (29) studied the retention, distribution and dosimetry of injected [ 224 Ra]RaCl 2 in six young adult beagles which were killed 0.04 to 7 days after [ 224 Ra]RaCl 2 administration. Their results suggest that, for the beagles, a fraction of roughly 0.08 of 220 Rn or 216 Po is produced in vivo and escapes from the skeleton. Increased in-vivo emanation of 220 Rn was not observed in a study by Klemm et al. (30) (32). Although a different set-up -diffusing alpha-emitters radiation therapy utilizing implantable sources carrying small activities of 224 Rathe arguments are applicable also to the case of bone metastases taking up 224 Ra. The released atoms disperse inside the tumor by diffusive and convective processes, creating, through their alpha emissions, a high-dose region measuring several millimeter in diameter about each source. If the decay point of 220 Rn is effectively the starting point for the migration of 212 Pb which may further distribute away from the source, the assessment by Arazi et al. (31) and Arazi (32) demonstrates that the size of the region subject to alpha particle irradiation may be expected to be of the order of millimeters rather than a few dozen micrometers. This might lead to a more homogeneous dose distribution in the tumor as compared to 223 Ra. Similar findings have been reported by Napoli et al. in an experimental study with 224 Ralabeled CaCO 3 microparticles (33). These considerations are not taken into account in any of the absorbed dose calculations until today (8,10,16 (18) showed radiation dose-related effects on chromosomal aberrations in peripheral lymphocytes after repeated treatments. The frequency of chromosomal aberrations observed during the course of therapy was related to the absorbed dose to the blood. They also observed, that the frequency of dicentric chromosomes induced in vivo agreed well with the corresponding value of dicentrics induced in vitro (18 (14). Concerning long-term side effects, Priest et al. (36) compared, in a reanalysis of the AS patient data of the Wick study, the higher incidence of radiation-induced cancer with the fact that the patient treatment resulted decreased pain and increased mobility. Both of which are associated with decreased mortality by non-cancer diseases and from all causes of death. In their analysis they found no excess mortality in the group of AS patients. According to the authors, "the study demonstrates the need to consider all causes of death and longevity when assessing health impacts following irradiation" (36).
With respect to long-term effects of treatment with [ 223 Ra]RaCl 2 , stochastic radiation-induced side-effects, although observed for [ 224 Ra]RaCl 2 , are less relevant in the context of cancer treatment of prostate cancer as the median survival time of patients after treatment is 14 months (1). This is significantly less than 2 years considered to be the latent period for induced leukemia or the 8 year average latent period for induced bone cancer (23,37,38). Therefore, presently the benefit of the treatment of prostate cancer patients with [ 223 Ra]RaCl 2 outweighs the hypothetical risk associated with this treatment.

Conclusions
When comparing the dosimetry data obtained by modelbased calculations on [ 223 Ra]RaCl 2 and [ 224 Ra]RaCl 2 or data obtained by bio-dosimetric methods no major differences are observed for most organs. For kidneys, liver and blood the differences, most likely, can be explained by the differing halflives of the respective progenies. Due to the difficulties associated with quantitative imaging of radium isotopes, absorbed doses derived by imaging procedures are less reliable due to inherent difficulties of image quantification. Furthermore, in vivo diffusion by radium progeny particularly in tumors is not well characterized and might need further experimental verification.
Data on long-term radiation-associated side effects are only available for treatment with [ 224 Ra]RaCl 2 . In several studies, an increased cancer mortality by leukemia and solid cancers was observed. Similar considerations likely apply to [ 223 Ra]RaCl 2 as the biokinetics and the absorbed doses are in the same range, but this increased risk may not be observed due to the much shorter life expectancy of prostate cancer patients treated with [ 223 Ra]RaCl 2 .