A Critical Review of Alpha Radionuclide Therapy—How to Deal with Recoiling Daughters?

Abstract

This review presents an overview of the successes and challenges currently faced in alpha radionuclide therapy. Alpha particles have an advantage in killing tumour cells as compared to beta or gamma radiation due to their short penetration depth and high linear energy transfer (LET). Touching briefly on the clinical successes of radionuclides emitting only one alpha particle, the main focus of this article lies on those alpha-emitting radionuclides with multiple alpha-emitting daughters in their decay chain. While having the advantage of longer half-lives, the recoiled daughters of radionuclides like ²²⁴Ra (radium), ²²³Ra, and ²²⁵Ac (actinium) can do significant damage to healthy tissue when not retained at the tumour site. Three different approaches to deal with this problem are discussed: encapsulation in a nano-carrier, fast uptake of the alpha emitting radionuclides in tumour cells, and local administration. Each approach has been shown to have its advantages and disadvantages, but when larger activities need to be used clinically, nano-carriers appear to be the most promising solution for reducing toxic effects, provided there is no accumulation in healthy tissue.

1. Alpha Emitting Radionuclides in the Clinic

Alpha-emitting radionuclides are a very promising type of radiotherapeutic agents showing great potential in the treatment of a broad range of malignancies [1,2,3,4,5,6,7,8,9,10,11]. Due to the short penetration depth of alpha particles, they are capable of destroying tumours while causing very limited damage to the surrounding healthy tissue. In fact, an emitted alpha particle will not travel much further than about 6 cell diameters (i.e., 50–100 µm). On the other hand, their high linear energy transfer (LET) gives them an increased relative biological effectiveness (RBE) [12] as compared to other radionuclide therapies. Just a few alpha particles through a cell nucleus are sufficient to destroy a cell, as cell death due to alpha-radiation is largely independent of oxygenation or active cell proliferation [13]. Furthermore, when alpha-emitting radionuclides are targeted to specific tumour cells in the body, they can be very effective in destroying metastases, which are difficult to treat by currently employed techniques.

Alpha radionuclide therapy has recently been introduced in the clinic, where it has seen a number of successes. Several clinical trials have been done with ²¹³Bi (bismuth), ²²¹At (astatine) and ²¹²Pb (lead) [1], and pre-clinical studies are being carried out with ¹⁴⁹Tb (terbium) [14]. Using tumour-specific monoclonal antibodies, ²¹³Bi has shown promise in the treatment of metastatic melanoma [2,3] and produced remissions in patients with acute myeloid leukemia [4,5,6]. A phase I study was completed on the use of locally injected ²¹³Bi-DOTA-substance P in the treatment of small gliomas. The patients received 1.07–2.00 GBq in one cycle or a total of 7.36 GBq in four cycles of ²¹³Bi-DOTA-substance P, of which more than 96% was retained at the tumour site. Most benefit has been seen in relatively small tumours, as sufficient intratumoural distribution has been achieved causing the tumours to be completely necrotic demarcated making it possible for non-operable gliomas to be surgically removed [7]. ²²¹At is also proving to be a promising alpha-emitter. A pilot study amongst patients with recurrent malignant brain tumours has shown the feasibility of the administration of ²¹¹At-ch81C6 into the surgically created resection cavity of patients, and has demonstrated minimal toxicity [8]. Women with recurrent ovarian carcinomas have been treated with varying therapeutic doses (33.6–119.2 MBq) of ²¹¹At bound to antibody fragments without detectable adverse effects, indicating that therapeutic doses can be delivered to micrometastases by the intraperitoneal administration of the ²¹¹At-MX35 F(ab’)₂ [9].

Despite of the clear promise of alpha radionuclide therapy, its use is limited to easily accessible tumours due the short half-lives of the applied isotopes: from 45.6 min for ²¹³Bi up to 7.2 h for ²¹¹At. Less easily accessible sites, where the targeting agent is taken up slowly, or solid tumours where a longer penetration time of the radiopharmaceutical is necessary, will need longer-lived alpha emitting radionuclides for optimal treatment [15]. So far, three long-lived alpha-emitters, namely ²²⁴Ra, ²²³Ra, and ²²⁵Ac, have made their appearance in the clinic. One of the first therapeutic uses of alpha-emitting radionuclides has been the intravenous injection of ²²⁴Ra-chloride in ankylosing spondylitis (AS) patients (a chronic inflammatory rheumatic disease). Between 1948 and 1975, 1588 patients were treated with repeated injections, receiving an activity of approximately 50 MBq. Amongst these patients, the cause of death could be determined for 1006. Although at the time positive therapeutic effects were confirmed for the AS patients, many patients, especially those aged below 21 when treated, developed malignant bone tumours. This increase in myeloid leukaemia as compared to the control group could be explained by the deposition of ²²⁴Ra into the bone. An increase in kidney and thyroid cancer was also observed, as compared to a control group [16,17]. In Germany, the pure ²²⁴Ra- chloride compound was approved again for the intravenous administration in AS patients in 2000, under the name ²²⁴SpondylAT^® [18]. The dose was much lower than what was used before, as patients now received total activities of only 10 MBq, with 1 MBq per injection [12]. Though the ²²⁴Ra treatment seemed to reduce medical costs and lost productivity [19], they could not deliver the clinical evaluation on time so it was discontinued in 2005 [20,21]. Since then ²²⁴Ra has not been used in clinical settings.

A comprehensive review has recently been published by Jadvar and Quinn [11] on the clinical trials of ²²³Ra dichloride (Xofigo^®, formerly Alpharadin^®) in the treatment of bone metastasis following castration-resistant prostate cancer. A short summary will be provided here for completeness. From 2008 to 2011 a total of 921 patients from 19 countries participated in a double-blind randomized phase 3 trial (ALSYMPCA). The patients in the treatment arm received six injections of 50 kBq/kg ²²³Ra, the patients in the placebo arm six injections with saline. A 3.6-month prolonged survival as compared to the placebo group has been observed [22]. On the completion of the trial, the treatment became FDA-approved. Dosimetric considerations following the intravenous injection of ²²³Ra dichloride report a dose of 16 Gy to the bone endosteum, and 1.5 Gy to the red bone marrow, while other organs received a lower dose [23]. Biodistribution experiments have shown that less than 2% of the daughter nuclides migrate away from the bone surface [24] Currently, Phase II trials are on-going for the use of ²²³Ra dichloride to help control breast cancer that has spread to the bones of breast cancer patients [25,26], although to date no results of these studies have been published.

The third radionuclide with alpha-emitting daughters that made it to the clinic is ²²⁵Ac. ²²⁵Ac-Labeled Humanized Anti-CD33 Monoclonal Antibody HuM195 (Actimab-A) has been used to treat patients with advanced myeloid malignancies. A Phase I trial has demonstrated that it is safe to use at doses ≤ 0.1 MBq/kg. Although dose-limiting toxicity (DLT) has occurred in one patient receiving 0.1 MBq/kg and two patients receiving 0.15 MBq/kg [27], blast reductions of more than 50% have been observed in 6 out of 12 patients [28]. No acute toxicities have been seen other than transient grade 2/3 liver function abnormalities, and no apparent damage to kidneys has been observed [10]. In a subsequent Phase-I trial, elderly patients with a total administered activity between 2.5 and 7.4 MBq have been treated. Bone marrow blast reductions have been seen in four out of six patients, but unfortunately no complete remissions have been reported. DLT has occurred in one patient with grade-4 thrombocytopenia, and other toxicities included grade 3 febrile neutropenia (n = 5) and pneumonia (n = 1). The FDA has cleared the compound for subsequent Phase I/II trials, where response rate, progression-free survival and overall survival will be assessed [29].

There is potentially a future for radionuclides with multiple alpha-emitting daughters like ²²⁵Ac (t_1/2 = 10.0 days) or ²²³Ra (t_1/2 = 11.4 days). However, toxicity is clearly still an issue due to the recoil energy the daughters experience upon alpha decay. This energy is in most cases at least 100 keV, more than 1000 times larger than the binding energy of any chemical compound. This simply means that bond rupture will always occur subsequent to alpha decay, implying that the released daughters, which are often themselves alpha emitters, might cause considerable harm (as seen in some of the clinical trials with e.g. AS patients [17]) since it will no longer be bound to a targeting moiety (Figure 1). Considering the presence of different proteins in blood and their high binding capacity the chance that the radionuclide will re-associate back with its original ligand is rather small. Therefore, in this review we will discuss the different alpha-emitting radionuclides and the strategies to limit their toxicity.

Figure 1. Schematic representation of a recoiling daughter radionuclide detaching from a targeting agent as a consequence of alpha decay.

Figure 1. Schematic representation of a recoiling daughter radionuclide detaching from a targeting agent as a consequence of alpha decay.

2. In Vivo Alpha Generator Radionuclides and Their Decay Chains

In Table 1 we have listed the potential alpha-emitting radionuclides having successive alpha-emitting daughters together with their decay characteristics and the recoil energy that the daughter radionuclides receive after decay. In several decay chains, multiple radionuclides can be used as in vivo generators for alpha-radionuclide therapy. These radionuclides are presented in bold-face in Table 1.

Table 1. Alpha radionuclides that can be used as in vivo generators and their decay characteristics [30].

Radionuclides and Their Decay Chain	Half-Life	Decay	Eα (MeV) *	Calculated ER (keV) *
225Ac	9.9 days **	α(100%)	5.8	-
221Fr	4.9 min	α(100%)	6.3	105.5
217At	32 ms	α(99.98%)/β−(0.01%)	7.1	116.9
213Bi	45.6 min	α(2%)/β−(98%)	-	132.8
213Po	4.2 µs	α(100%)	8.4	-
209Pb	3.3 h	β− (100%)	-	160.4
209Bi	stable	-	-	-
227Th	18.7 days	α(100%)	6	-
223Ra	11.4 days	α(100%)	5.7	108.4
219Rn	4.0 s	α(100%)	6.8	104.5
215Po	1.8 ms	α(100%)	7.4	126.9
211Bi	2.2 min	α(99.7%)/β−(0.3%)	6.6	140.1
207Tl	4.8 min	β− (100%)	-	128.1
207Pb	stable		-	-
228Th	1.9 years	α(100%)	5.4	-
224Ra	3.7 days	α(100%)	5.7	96.9
220Rn	55.6 s	α(100%)	6.3	103.4
216Po	0.15 s	α(100%)	6.8	116.5
212Pb	10.6 h	β−(100%)	-	128
212Bi	60.6 min	α(36%)/β−(64%)	6.1/-	-
208Tl/212Po	3.0 min/0.3 µs	β−(100%)/α(100%)	-/8.8	116.5/-
208Pb	stable	-	-	-
230U	20.8 days	α(100%)	5.9	-
226Th	31 min	α(100%)	6.3	104.3
222Ra	38 s	α(100%)	6.5	114.2
218Rn	35 ms	α(100%)	7.1	120.4
214Po	164 µs	α(100%)	7.7	133.4
210Pb	22.3 years	β−(100%)		146.5
210Bi	5.0 days	β−(100%)		-
210Po	138.4 days	α(100%)	5.3	-
206Pb	stable			-

* The alpha and recoil energies have been rounded up;** this value is from Pommé et al. [31].

Table 1. Alpha radionuclides that can be used as in vivo generators and their decay characteristics [30].

Radionuclides and Their Decay Chain	Half-Life	Decay	Eα (MeV) *	Calculated ER (keV) *
225Ac	9.9 days **	α(100%)	5.8	-
221Fr	4.9 min	α(100%)	6.3	105.5
217At	32 ms	α(99.98%)/β−(0.01%)	7.1	116.9
213Bi	45.6 min	α(2%)/β−(98%)	-	132.8
213Po	4.2 µs	α(100%)	8.4	-
209Pb	3.3 h	β− (100%)	-	160.4
209Bi	stable	-	-	-
227Th	18.7 days	α(100%)	6	-
223Ra	11.4 days	α(100%)	5.7	108.4
219Rn	4.0 s	α(100%)	6.8	104.5
215Po	1.8 ms	α(100%)	7.4	126.9
211Bi	2.2 min	α(99.7%)/β−(0.3%)	6.6	140.1
207Tl	4.8 min	β− (100%)	-	128.1
207Pb	stable		-	-
228Th	1.9 years	α(100%)	5.4	-
224Ra	3.7 days	α(100%)	5.7	96.9
220Rn	55.6 s	α(100%)	6.3	103.4
216Po	0.15 s	α(100%)	6.8	116.5
212Pb	10.6 h	β−(100%)	-	128
212Bi	60.6 min	α(36%)/β−(64%)	6.1/-	-
208Tl/212Po	3.0 min/0.3 µs	β−(100%)/α(100%)	-/8.8	116.5/-
208Pb	stable	-	-	-
230U	20.8 days	α(100%)	5.9	-
226Th	31 min	α(100%)	6.3	104.3
222Ra	38 s	α(100%)	6.5	114.2
218Rn	35 ms	α(100%)	7.1	120.4
214Po	164 µs	α(100%)	7.7	133.4
210Pb	22.3 years	β−(100%)		146.5
210Bi	5.0 days	β−(100%)		-
210Po	138.4 days	α(100%)	5.3	-
206Pb	stable			-

* The alpha and recoil energies have been rounded up;** this value is from Pommé et al. [31].

²²⁸Th (thorium) and ²²⁴Ra are the products of two decay chains: The one of ²³²Th (²³²Th$\stackrel{\alpha }{\to }$²²⁸Ra$\stackrel{\beta -}{\to }$²²⁸Ac$\stackrel{\beta -}{\to }$²²⁸Th) and the decay of ²³²U (uranium) (²³²U$\stackrel{\alpha }{\to }$²²⁸Th). Starting from ²³²Th, ²²⁸Ra is extracted and left to decay to ²²⁸Th, which is then applied to prepare a ²²⁴Ra generator. ²²⁸Th can also be obtained by successive neutron capture of ²²⁶Ra, resulting in ²²⁸Ra which decays in two beta decay steps to ²²⁸Th. ²²⁸Th, due to its long half-life, is only used to generate ²²⁴Ra. This decay series produces four alpha particles, counted from ²²⁴Ra, and four recoiling daughters of which ²²⁰Rn, ²¹²Pb and ²⁰⁸Tl (thallium) are the most important ones in terms of half-life. In addition to the recoils released by the alpha decay, ²¹²Bi, which is also an alpha emitter, can be freed due to the internal conversion of γ-rays emitted by the excited ²¹²Bi nuclides. Thus, in this decay series there are actually five daughters that might be harmful [32].

²²⁷Th and ²²³Ra are both decay products of ²²⁷Ac, which is extracted from the mill waste of ²³⁵U mines. ²²⁷Th can be separated from its mother using anion exchange chromatography [33], while ²²³Ra is typically supplied from a ²²⁷Ac generator containing an actinide chromatographic resin (Dipex 2) [34]. This decay chain generates five alpha particles and recoiling daughters of which ²²³Ra, ²¹¹Pb, and ²⁰⁷Tl, are the most essential. Radium is in this case the most hazardous one due to its considerably long half-life and the fact that it is part of a decay chain having four alpha particles. This is also the reason to directly use ²²³Ra for treatment rather than ²²⁷Th, although it comes with the additional challenge of finding a chelate able to make a stable complex with radium. In addition, the increased demand for ²²³Ra will require the development of alternative production routes of ²²⁷Th.

²²⁵Ac and ²¹³Bi are typically extracted from ²²⁹Th, originating from ²³³U, produced by neutron capture of ²³²Th (²³³Th$\stackrel{\beta -}{\to }$²³³Pa$\stackrel{\beta -}{\to }$²³³U$\stackrel{\alpha }{\to }$²²⁹Th). Due to increasing demand of ²²⁵Ac, shortage via this production route is expected and alternative methods are being investigated, from which the proton irradiation of ²²⁶Ra targets appears to be the most promising process [35]. An alternative to this route is the bombardment of ²³²Th targets by protons having energy of less than 200 MeV, as researched by Weidner et al. [36]. The ²²⁵Ac chain has four major alpha particles and recoiling daughters, from which ²²¹Fr, ²¹³Bi and ²⁰⁹Pb are long-lived enough to induce undesired damage due to subsequent alpha or beta emission.

²³⁰U is produced in cyclotrons typically by the proton bombardment of ²³²Th, resulting in ²³⁰Pa which decays to ²³⁰U, allowing the extraction of the carrier-free product [35]. The ²³⁰U chain has five alpha particles before reaching the long-lived beta emitter ²¹⁰Pb. There are in this case only two recoil daughters of concern²²⁶Th and ²²²Ra, if we do not count ²¹⁰Pb, a. An additional problem here, similar to radium, is the availability of appropriate chelates allowing in vivo applications with little loss of uranium.

3. Distribution of Recoil Daughters in the Body

There are three processes that contribute to the distribution of the recoiling daughters: (1) distance covered due to recoil energy, (2) diffusion processes and (3) active transport such as convection or blood flow. The importance of each process will depend on the location where the recoil atoms are released. The instantaneous energy that a recoil atom gets will make it cover on average around 100 nm and is therefore mostly responsible for the breaking of chemical bonds or the possible escape from carriers [37]. It is the indirect process that allows the recoils to acquire different speciation and distribution than originally intended. Diffusion transport becomes important once the radionuclide reaches certain tissues or organs, such as tumours or kidneys for instance, and will be dependent on the medium. It will thus differ between diffusion in blood-like medium and in the extra/intracellular matrix, as it would be subject to the interaction of the ions with blood or cell components. Most of the time the recoils will be released in the blood stream and their eventual fate will be determined by their affinity for certain organs/tissues provided that they live long enough to reach their biological destination. Table 2 shows a summary of the target organs of several of the elements associated with recoiling daughters.

Table 2. Major targeted organs based on 24 h distribution after intravenous injection (IV).

Element	Major Targeted Organs *
Francium	primarily kidneys [38]
Bismuth	30% urine, 40% kidney, 30% other organs [39]
Radium	25% bone surface, 45% soft tissue, 30% excreted via large intestine [12,24]
Radon	soft tissue to blood: 100 day−1, exhaled 1 min−1, bone to blood: 0.36 d−1 [40]
Lead	55% blood, 15% liver, 10%–15% skeleton, kidneys 4% after 1 day [40]
Polonium	28% liver, 28% kidneys, 10% red bone marrow, 5% spleen [41]

Table 2. Major targeted organs based on 24 h distribution after intravenous injection (IV).

Element	Major Targeted Organs *
Francium	primarily kidneys [38]
Bismuth	30% urine, 40% kidney, 30% other organs [39]
Radium	25% bone surface, 45% soft tissue, 30% excreted via large intestine [12,24]
Radon	soft tissue to blood: 100 day−1, exhaled 1 min−1, bone to blood: 0.36 d−1 [40]
Lead	55% blood, 15% liver, 10%–15% skeleton, kidneys 4% after 1 day [40]
Polonium	28% liver, 28% kidneys, 10% red bone marrow, 5% spleen [41]

Several studies have been dedicated for establishing the fate of alpha emitters, including mother and daughter ions. For instance, the radiotoxicity of ²²⁵Ac has been evaluated in mice. Biodistribution studies have been done with ²²⁵Ac chelated to four different compounds: cyclohexyl diethylenetriamine pentaacetic acid (CHX-DTPA), 1,4,7,10,13-pentaazacyclopentadecane-N,N′,N′′,N′′′,N′′′′-pentaacetic acid (PEPA), ethylene diamine tetraacetic acid (EDTA) and acetate. In this work it has been found that ²²⁵Ac mainly goes to the liver, but also accumulates in the femur, kidneys and heart, and although its concentration in the liver and bone increases with time, it is cleared from the kidney and the heart. The ²²⁵Ac-acetate complex has had a 5.5 times higher %ID/g in the liver than the ²²⁵Ac-CHX-DTPA and the ²²⁵Ac-PEPA complexes. It must be noted that the biodistribution of ²²⁵Ac has been determined immediately after sacrifice using the 218 keV gamma emitted by ²²¹Fr, although with its 4 min half-life ²²¹Fr could have recoiled and diffused away from the location of the mother isotope [42]. A different study by G.J. Beyer et al. has also reported a high liver uptake of ²²⁵Ac in mice and rats [43].

Radium naturally targets the hydroxyapatite matrix in the bone and is thus the ideal agent for the treatment of bone metastasis. The rapid cascade of alpha particle emitting daughters will thus deliver high doses to bone tumours. Although the daughter nuclides are not intrinsically bone-seeking, their short half-life appears to prevent them from doing major damage to healthy tissue. An in vivo study has shown that less than 2% of the daughters migrate away from the bone surface within 6 h after administration of ²²³Ra, and after three days this number has dropped down to less than 1% [24]. This effect is due to the long half-life of ²²³Ra, such that after a few days it has been incorporated better into the bone tissue. The retention of the daughters in bone thus appears similar to that of the parent nuclide. However, when ²²³Ra would be targeted to other sites than bone tumours, recoiling daughters would be more likely to escape from the targeted site and distribute according to their biological affinity, once again depending on half-life and location.

To learn more about the effects observed in the group of AS patients treated with ²²⁴Ra as described earlier, Lloyd et al. have attempted to acquire more information about the biodistribution of ²²⁴Ra and its daughters [44]. In their study, six young adult beagles were killed from 1 h to 7 days after a 3.4 MBq ²²⁴Ra injection, and activities of ²²⁴Ra, ²¹²Pb, and ²¹²Bi were counted in humerus, lumbular vertebrae, eyes, kidneys, and the liver. Biological retention of the ²²⁴Ra in the bones ranges from 37% at 1 h till 58% after 8 h, decreasing slowly to 42% after a week. In the soft tissue on the other hand, 61% of the initial activity has been found 1 h after administration, diminishing to only 6% at 7 days. About 8% of the ²²⁰Rn gas has left the body. Furthermore, substantial amounts of ²¹²Bi and ²¹²Pb have been found in red blood cells. The kidneys contained mainly ²¹²Bi, and the liver had excess amounts of ²¹²Pb (20% of total) as compared to ²²⁴Ra. The eyes and bones contained relatively less ²¹²Bi and ²¹²Pb than ²²⁴Ra. These results indicate significant redistribution of the daughter nuclides. This higher redistribution as compared to that found in ²²³Ra (t_1/2 = 11.4 days) studies can be partially explained by the shorter half-life of ²²⁴Ra (t_1/2 = 3.7 days), as the mother isotope does not have enough time to move deep into the bone surface for the daughters to be also retained in the bone. In addition, the gaseous daughter ²²⁰Rn (t_1/2 = 55 s) has a much longer time to redistribute itself in the body as compared to ²¹⁹Rn (t_1/2 = 3.9 s), the ²²³Ra daughter. These are the two main reasons that ²²⁴Ra therapy has shown adverse effects in the kidneys, eyes and livers of treated patients [44].

4. Approaches to Deal with the Recoil Problem

There are generally speaking three strategies proposed in literature to deal with recoil issues: encapsulation in a nano-carrier, fast uptake of the alpha emitters in tumour cells, and local administration such as intratumoural injection.

4.1. Encapsulation in A Nano-Carrier

The first approach is to apply some kind of a nano-carrier that is capable of retaining the recoiling daughters. So far zeolites, liposomes, polymersomes, and metal-based particles have been investigated. The actual recoil retention is difficult to be compared though, as determining the percentage of recoiling daughters is not always straightforward, especially in systems where re-association is possible. For instance, Piotrowska et al. have used zeolites as carriers for ²²⁴Ra, showing that the percentage of recoiled daughters (²¹²Bi, ²¹²Pb and ²⁰⁸Tl) escaping from the zeolites is relatively small even under blood serum conditions [45]. However, it needs to be mentioned that in these experiments, most likely, equilibrium has been established and the determined values in fact reflect the distribution of the different ions according to their stability constants. Once injected in the blood stream, the ejected radionuclides will not be in thermodynamic equilibrium anymore and their distribution might be of a very different kind.

Sofou et al. have used different types of liposomes (zwitterionic, cationic and anionic) to perform retention studies of ²²⁵Ac and its daughters. ²¹³Bi retention has been revealed to have weak size dependence for liposomes sized between 100 and 800 nm a maximum of 12% remained in the carriers [46]. At the same time, Wang et al., have investigated polymeric vesicles for the same decay chain and have reported that nearly 70% of ²¹³Bi is still present in carriers of 800 nm [47]. These discrepancies can be due to the relatively fast decay of ²²⁵Ac that creates experimental difficulties prone to measurement errors strongly dependent on the separation method. In the case of ²²⁵Ac decay, the separation method should be fast enough otherwise the initial activity of ²²¹Fr and ²¹³Bi might be wrongly calculated, resulting in underestimation of the recoil retention. Typically after separation the time is monitored accurately and the activity of ²²¹Fr and ²¹³Bi is determined and plotted against time. The initial activity is then estimated using a simplified version of the secular equilibrium decay equation: $A_2\left(t\right)=\frac{{\text{λ}}_2}{{\text{λ}}_2-{\text{λ}}_1}{A}_1\left(0\right)\left(e^{-{\text{λ}}_{1}t}-e^{-{\text{λ}}_{2}t}\right)+A_2\left(0\right)e^{-{\text{λ}}_{2}t}$, where A₁ and A₂ reflect the activity of the mother and daughter respectively, while λ₁ and λ₂ correspond to the mother and daughter decay constants. A slow separation does not allow for determining the ²²¹Fr escape and leads to uncertainties of the initial ²¹³Bi concentration, i.e., the nearer the measured time is to time zero the more accurate the determination of the initial activity and hence the retention.

Liposomes of 80 nm in size have also been investigated as carriers of ²²³Ra, but unfortunately all available recoil data is based on in vivo experiments, which complicates determining the actual retention. Nevertheless, from this study is clear that that the recoil daughters ²¹¹Pb and ²¹¹Bi (the third and fourth daughter of ²²³Ra) are not retained sufficiently in the liposomes, as the activity measured in the kidneys has been found to be much higher than the expected value based on equilibrium assumptions, calculated from the ²²³Ra activity in the same organ. In addition to this ²²³Ra has also been measured in the bones of the lab-animals. These findings strongly suggest that relatively small sizes of this type of soft nano-carriers will not retain a sufficient percentage of recoils, which can also be anticipated from range calculation of recoiling atoms based on their stopping power [48]. In soft materials like polymers, for instance, the projected range of ²²¹Fr has been reported to be around 90 nm, which can already be larger than the applied nano-carrier [37]. An alpha cascade will result in an even larger distance that the recoils can cover. Therefore in order to achieve sufficient recoil retention materials having higher Z are necessary. This line of reasoning has been followed by Woodward et al. and McLaughlin et al., both working on lanthanide-based phosphate nanoparticles doped with ²²⁵Ac, except the latter has also applied gold coating to the carriers [49,50]. The retention of ²²¹Fr has been determined in vitro, while the fate of ²¹³Bi has also been assessed in vivo. The gold-layered particles have given the most promising results with nearly 100% initial retention of ²²¹Fr when four layers of gadolinium phosphate are applied, but unfortunately these particles could not be used in vivo due to sedimentation issues. Biodistribution studies based on ²²⁵Ac-nanoparticles having less than four gadolinium phosphate layers coupled to mAb 201b have revealed a ²¹³Bi retention of at least 69% and a much-decreased renal uptake of free ²¹³Bi. Although these results are very promising, there is still liver uptake of the ²²⁵Ac doped particles due to their size. Considering the nature of the particles it is expected that they will remain in the liver for a considerable time, which might cause adverse effects and should be studied further. Smaller particles may prove to have better bio-distribution and faster clearance but on the expense of recoil retention. Clearly an optimum between recoil retention and uptake of nano-carriers by healthy organs needs to be determined to make this approach applicable in the clinic.

4.2. Fast Uptake in Tumour Cells

The second approach is to ensure that the radiopharmaceutical is rapidly taken up by the tumour cells and that the remaining not-adsorbed part is excreted fast from the body. This argument has been used for both peptides and antibodies. Since antibodies take a long time to accumulate at the tumour site, one can imagine that smaller targeting agents would fit better to this strategy. Nevertheless, this approach is almost entirely applied with larger targeting agents. Similarly in the case of ²²³Ra and its daughters, their affinity to bone can be used to limit the toxicity to other tissues.

²²⁵Ac has been investigated in conjugation with vascular tumour-homing peptide F3 [51] and with octreotide [52] in mice resulting in a rather different kidney toxicity. In the case of the F3 peptide little differences have been observed between ²²⁵Ac and ²¹³Bi concerning both therapeutic efficiency and kidney toxicity. The octreotide results have revealed better efficiencies when compared to ¹⁷⁷Lu in the treatment of neuroendocrine tumours, but tubular necrosis at activities above 30 kBq has been reported. The biodistribution in this case has shown high tumour uptake within 30 minutes and absence of blood activity within 4 h after injection. The kidney uptake, although rapidly decaying since the injection time point, has been found to be still above 5%ID/g at 24 h [52]. No biodistribution data has been provided in the F3 peptide study. An explanation for these toxicity differences can be the rather larger activities of ²²⁵Ac applied with octreotide, i.e., 30 kBq vs. 1.85 kBq used in the F3 peptide conjugation. In addition to this the mice in the F3 peptide studies have been followed only for 3 months unlike in the octreotide experiments where a period of 7 months has been chosen, which implies that any retarded effects could simply not have been detected in the first case.

Significantly more attention has been paid to the use of antibodies in alpha radionuclide therapy. One of the first pre-clinical studies using ²²⁵Ac has been carried out by McDevitt et al. who aimed at a high degree of tumour cell internalization as a solution to the recoil problem [53]. In vivo experiments have been performed using two different animal models, prostate carcinoma and disseminated lymphoma, and respectively antibodies J591 and B4 conjugated to DOTA. In addition to this in the first model the actual retention of ²²¹Fr and ²¹³Bi has been estimated in the tumour tissue, revealing the presence of 88% of ²²¹Fr and 89% of ²¹³Bi of the ²²⁵Ac secular equilibrium values. The results of these studies indicated a clear benefit of alpha radiation at low administered activities with no toxic effects. Further in this investigation the in vitro stability of Ac-DOTA bound to lintuzumab (HuM195) has been determined revealing little dissociation of ²²⁵Ac. The same antibody coupled to ²²⁵Ac has also been applied in experiments concerning non-human primates. The half-life blood circulation time of the antibody has been found to be 12 days with 45% blood clearance of ²¹³Bi. In addition to this at low activities (28 kBq/kg) no toxicity has been observed within 6 months, but at cumulative doses of 215 to 370 kBq severe toxic effects become apparent. These toxic effects are believed to be caused by the released daughters from the unbound antibody, ²²¹Fr and ²¹³Bi, which are expected to accumulate in the kidneys [54]. These results suggest that remaining below activities of 28 kBq/kg should not lead to adverse effects and the combination of this antibody with ²²⁵Ac has actually been translated to clinical studies as described earlier in this paper [55].

²²⁵Ac in combination with trastuzumab, an anti-HER2/neu antibody, has also been tested in ovarian cancer mice models applying intraperitoneal injection and three levels of activities: 8.1, 12.2 and 16.7 kBq. The results suggest that fractionated therapy using less activity might be more advantageous than a single injection of the same total activity based on early mortality numbers and median survival time, which appeared to be optimal when administering less activity. The treatment has been concluded to be successful because of the reached significant extension of the survival time and limited toxicity [56].

²²⁵Ac coupled to anti rat HER-2/neu monoclonal antibody (7.16.4) has been studied in the treatment of breast cancer metastases. The therapeutic results have shown remarkable prolongation of mice survival time, i.e., up to 1 year, and better efficiency than ²¹³Bi or ⁹⁰Y controls. However, the biodistribution findings have indicated serious damage of the kidneys, probably caused by the two daughters of the alpha emitter having the longest half-life, i.e., ²²¹Fr and ²¹³Bi [15].

The success of therapy with ²²⁵Ac will depend on the extend to which renal toxicity can be reduced. In principle, renal damage can be greatly diminished by means of administering a metal scavenger and a diuretic as reported by Jaggi et al. [38]. In this case the metal chelate (2,3-dimercapto-1-propanesulfonic acid, DMPS) administered orally has significantly suppressed ²¹³Bi kidney uptake by, most likely, complexing with the metal and prolonging its circulation time in the blood. On the other hand, diuretics such as furosemide and chlorothiazide have been used to inhibit the absorption of ²²¹Fr in renal tubular cells showing that excretion is stimulated and kidney damage is reduced.

Besides ²²⁵Ac, ²²⁷Th conjugated to the monoclonal antibody DAB-4 has been evaluated in combination with chemotherapy in mice with induced Lewis lung tumours. Mice have had better median survival times even when ²²⁷Th has been applied on its own with activities of 20 kBq, but the therapeutic efficiency has been shown to be more pronounced when combined with chemotherapeutics. Adverse health effects have not been observed but considering the very short survival times of the mice, (i.e., in the best case 19 days), this is not surprising and the authors have acknowledged that bone uptake might become a problem in malignant cases having longer life expectancy [57]. Indeed the development of secondary cancers might be expected considering the fact that ²²³Ra is the daughter of ²²⁷Th and that it accumulates in bone but no data could be found in this particular publication to support these expectations. Further elaborate studies concerning the toxicity of ²²⁷Th coupled to rituximab have been carried out by Dahle et al. who investigated different activity levels and their effect on mouse body weight, white blood cells and platelet counts, and histological examination. The no-observed-adverse-effect-level has been found to be 200 kBq/kg, although the maximum tolerated activity has been estimated to be much higher, i.e., between 600 and 1000 kBq/kg. The contribution of ²²³Ra to bone marrow toxicity could not be properly evaluated due to cross-contaminations [33].

4.3. Local Administration

The third approach is to insert or inject the alpha-emitting radionuclides directly in the or near the tumour tissue, as has been tested in Phase I clinical studies with ²¹³Bi-DOTA-substance P locally injected in gliomas by Cordier et al. [7]. This approach is also used by Cooks et al, who have made clever use of the recoiling properties of alpha emitting radionuclides in a series of studies with ²²⁴Ra [58,59,60,61,62]. They have developed ²²⁴Ra-loaded wires that continuously release radioactive daughters ²²⁰Rn, ²¹⁶Po and ²¹²Pb from their surface, while the ²²⁴Ra itself remains in the wires. According to their theoretical studies, the region around the site of wire insertion affected by the diffusion of ²²⁰Rn, the recoiled daughter of ²²⁴Ra, is in the order of millimetres, reaching a dose of 10 Gy up to 2.6 mm away from the inserted wire at an initial ²²⁰Ra activity of 37 kBq. Experimental studies show that the ²²⁰Rn daughter ²¹²Pb travels substantially further away from the tumour. In 0.1 g tumours, up to about 90% of the ²¹²Pb leaks from the tumours, while for 2.4 g tumours this is only 12%. Most of the ²¹²Pb leaving the tumour is absorbed by the kidneys (8.5 ± 5.8%) [58]. Despite the high kidney dose, this new form of “brachytherapy”, called diffusing alpha-emitters radiation therapy (DaRT), has shown great promise in the treatment of both pancreatic [61], and lung carcinomas [59], as well as a number of other human-derived tumours [59,60]. Recently, an anti-tumour immunity in mice previously treated with DaRT therapy has been found, where 77% of the population did not develop tumours after re-inoculation of the tumour cells, 21 days after the treatment with the ²²⁴Ra wires (versus 33% in the control group). The survival of DaRT-treated mice has increased from 29% (untreated) to 63% [62], suggesting that improved long-term cancer immunity might arise from tumour antigens supplied by destroyed tumour cells. This form of therapy could prove very beneficial for patients with relatively large, accessible tumours, though for smaller metastasis different methods would need to be applied.

5. Conclusions

This paper has presented a review on alpha radionuclide therapy. During the past few years, some progress has been made in getting a few alpha radionuclide therapy concepts to the clinic, however, by and large we find that in most studies concerning long-lived alpha-emitters, recoiling daughter atoms pose a serious problem. If the distribution of recoiled atoms is not controlled, toxic effects due to the recoiling daughters will be present. At higher injected radionuclide activities, this could induce significant damage to healthy tissue. The question is whether the activity that can be injected without toxic effects will be sufficient to have the expected (much improved) therapeutic effects. The success of each approach depends to a large extent on the type of radionuclide (e.g., ²²³Ra) and the type of tumours to be treated. For instance, in systemic radionuclide therapy to treat metastasized tumours other than bone malignancies, nano-carriers that can limit the distribution of the recoiling daughters may offer a solution. In this case, the carriers that do not accumulate at the tumour site should also not be retained in healthy organs. Rapid clearance of these particles from the organs is thus essential. On the other hand, tumours that are easily accessible can profit from direct insertion of the alpha-emitter in a tumour volume. Alpha-radionuclide therapy is thus promising, provided that the recoil obstacle can be overcome first. Only then will patients have the full benefit of this treatment.