Decay scheme of the 223Ra series
4.1 Some Historical Remarks
The discovery of radium (L. Radius, ray) by Marie and Pierre Curie in 1898, represents a unique chapter in the history of natural sciences and opened a new era in the treatment of cancer (Harvie 1999; Mould 1999; Macklis 2002). The elemental form was isolated in 1910 by electrolysis of a solution of pure radium-chloride, employing a mercury cathode. After distillation in an atmosphere of hydrogen this amalgam yielded the pure metal. Originally, radium was obtained from the rich pitchblende ore found in Joachimsthal, Bohemia, with approximately 1 g of radium in 7 tonnes of mineral. Radium is the heaviest of the earth alkaline elements, and all isotopes are radioactive (Table 1). Radium-226 has the longest half-life (t1/2 = 1600 years), is the most abundant radium isotope in nature and is formed in the decay chain from 238U. Other naturally occurring radium isotopes are 228Ra and 224Ra from the 232Th-family and 223Ra from the 235U-family. 1 g of 226Ra undergoes 3.7 × 1010 disintegrations per s; a number defining the unit Curie.
Summary of half-lives, daughter nuclides and decay properties for 223Ra, 224Ra, 226Ra and 228Ra
223Ra (11.43 d, α)
224Ra (3.66 d, α)
226Ra (1600 y, α)
228Ra (5.75 y, β)
219Rn (3.96 s, α)
220Rn (56 s, α)
222Rn (3.8 d, α)
228Ac (6.13 h, β)
215Po (1.78 ms, α)
216Po (0.15 s, α)
218Po (3.05 min, α)
228Th (1.91 y, α)
211Pb (36.1 min, β)
212Pb (10.6 h, β)
214Pb (26.8 min, β)
224Ra (3.66 d, α)
211Bi (2.14 min, α)
212Bi (60.6 min, β)
214Bi (19.8 min, β)
220Rn (56 s, α)
207Tl (4.77 min, β)
212Po (0.4 μs, α)
214Po (162 μs, α)
216Po (0.15 s, α)
210Pb (22 y, β)
212Pb (10.6 h, β)
210Bi (5.0 min, β)
212Bi (60.6 min, β)
210Po(138 d, α)
212Po (0.4 μs, α)
An important question is to what extent radium is carcinogenic. It is well known that bone cancer may occur in subjects exposed to 226Ra (Martland and Humphries 1929). However, at lower dose levels (below 5 Gy to the bone for 226Ra) no significant increase in cancer was observed in humans (reviewed in Hall 2000). In the concept of treating patients with CRPC; so far not a curable situation, radiation risk should be put into perspective.
For several decades injections of 224Ra (t½ = 3.7 days) were used to treat ankylosing spondylitits (Tiepolt et al. 2001; Wick et al. 1999, 2009; Nekolla et al. 2000; Lassmann et al. 2002; Spiess 2010). It was reintroduced for this indication in Germany a few years ago (Tiepolt et al. 2001), but is to the authors knowledge today no longer available. In the case of 224Ra, long-term follow-up of adult patients receiving moderate levels revealed no significant difference in overall cancer incidence or life expectancy as compared to a control population (Wick et al. 1999, 2009; Nekolla et al. 2000; Lassmann et al. 2002; Spiess 2010).
The first report on use of cationic 223Ra in humans was published more than 50-years ago and was related to a biokinetic study with tracer amounts of radioactive alkaline earth metals for the purpose of acquiring data for radiation safety assessments (Norris et al. 1955). The first study with relevant quantities of dissolved 223Ra salt was initiated in 2001 (Henriksen et al. 2001). As with dissolved 89SrCl2 (Metastron), radium cations are incorporated within the bone matrix of metabolically active bone, probably by inclusion in the calcium phosphate and hydroxyapatite crystals (Bruland et al. 2006; Henriksen et al. 2003).
4.2 Technology Platform
Source material, 227Ac, was during initial development prepared from irradiated samples of 226Ra. By cation and anion exchange methods 227Ac was prepared in good purity from 226Ra (Henriksen et al. 2001). After allowing time for in-growth, 223Ra was separated from 227Ac and 227Th trapped in Ac-resin. The 223Ra eluted from the resin had high purity with no detectable amounts of the longer lived generator nuclides 227Ac and 227Th as measured by gamma spectroscopy on fresh and on decayed (>6-month old) samples (Henriksen et al. 2001). The 223Ra was dissolved in physiologically compatible sodium chloride/sodium citrate buffer and sterile filtered. The cationic integrity of 223Ra was determined using barium sulphate as a precipitation agent, yielding routinely better than 99 % precipitation.
5 Pre-clinical Experiments
The chloride formulation of 223Ra has been studied in pre-clinical models. In normal mice, the biodistribution of 223Ra was shown to correspond to that of 89Sr, with targeting of the skeleton, little soft tissue uptake with minimal translocalization and good retention of 223Ra and its daughter isotopes within the bone matrix (Henriksen et al. 2003). Modelling the dose deposition in relation to tumour deposits within the bone marrow suggested a significant reduction in dose to the normal bone marrow with 223Ra as compared to 89Sr (Henriksen et al. 2003). In a rat model of metastatic breast cancer, 223Ra showed significant anti-tumour effects in the absence of bone marrow toxicity (Henriksen et al. 2002). Animals treated with >100 kBq/kg 223Ra had 40 % survival beyond the 67-day follow-up period compared to 0 % in control animals. Treatment with conventional chemotherapeutics, a beta-emitting BSR or the bisphosphonate pamidronate gave no survival benefit in this animal model.
A study addressing acute and sub-acute toxicity in mice from injected 223Ra as dissolved salt, indicated that the bone marrow and the bone associated cells were severely affected at very large dosages of 223Ra, but that animals could survive doses far above those considered clinical relevant in humans in terms of kBq/kg (Larsen et al. 2006).
Thus, animal data and dosimetric studies indicated that this bone-targeted alpha-emitter could deliver therapeutically relevant radiation doses to bone surfaces and skeletal metastases at activity levels acceptable in terms of bone marrow radiation exposure.
6 Radium-223: Clinical Studies
A clinical development program for 223RaCl2 (Alpharadin™) was initiated based on the results presented above and upon approval obtained from the relevant institutional review boards and regulatory authorities.
6.1 Phase 1 Studies
In a single-dosage administration with increasing levels of 223Ra (from 46 to 250 kBq/kg) in 25 patients with bone metastases from breast and prostate cancer, dose-limiting haematological toxicity was not reached (Nilsson et al. 2005). Mild and reversible myelosuppression was observed in some cases, with only grade 1 toxicity for thrombocytes at the two highest dose levels. Quality of life was evaluated at baseline and at 1, 4 and 8 weeks after injection, and pain relief was indicated at these time points in more than 50 % of the patients (Nilsson et al. 2005). Notably, a decline in total serum alkaline phosphatase; a marker for abnormal bone metabolism in metastatic prostate cancer, greater than 50 %, was observed among patients with elevated pre-treatment values. Radium-223 showed rapid blood clearance with only 12 % of its initial value at 10 min after injection, and a further reduction to 6 % at 1 h and to less than 1 % at 24 h after infusion. Gamma-camera scintigraphy indicated that 223Ra accumulated in skeletal lesions similar to patterns observed in diagnostic bone scans with 99mTc-MDP. Radium-223 cleared mainly by the intestinal route, in accordance with findings in older studies with dogs and primates receiving injections with dissolved radium salts (Nilsson et al. 2005).
A small phase 1B feasibility study involving six patients with advanced prostate cancer was then performed (Bruland et al. 2006, 2008b) to evaluate the safety profile of repeated 223Ra injections. Six prostate cancer patients were administered up to 250 kBq kg−1 body weight, either as a fractionated regimen of two injections of 125 kBq kg−1 bodyweight with a 6-week interval (two patients) or 50 kBq kg−1 body weight given five times with a 3-week interval (four patients). The patients in the 50 kBq kg−1 × 5 group did not experience any additional toxic effects related to repeated treatment compared with the single-injection in the phase 1A study. The haematological profiles were less determined by the fractionation schedule compared to a single dosage totalling the same as the five fractions combined. Because of non-skeletal disease progression, only one of the patients in the 125 kBq kg−1 × 2 group actually got the second dosage. Of the two patients not given the 50 kBq kg−1 follow-up dosages, one died due to progression of liver metastases, and the other was deemed unfit for further treatment due to recurrence of a previous heart condition. Mild and reversible myelosuppression occurred, with nadir 2 to 3 weeks after injection, and complete recovery during the follow-up period. The thrombocytes revealed only grade 1 toxicity, whereas neutropenia of maximum grade 3 occurred in one of the patients. Experience from this small phase 1B study indicated that repeated administration of 223Ra was well tolerated, and that the time span between injections should be scheduled according to the dosages given, so that the blood cell count could normalize before a new injection.
6.2 Phase 2 Clinical Studies
BC 1–02. In this phase 2 randomized trial late stage prostate cancer patients scheduled to receive external beam radiation towards the dominating painful site, either received saline injections (four times with 4-week intervals) or four times repeated 223Ra (50 kBq/kg) given at 4-week intervals (Nilsson et al. 2007). Treatment with 223Ra resulted in a statistically significant decrease in bone alkaline phosphatase from baseline compared to placebo, showing the strongest decrease in patients with elevated pre-treatment levels. The median relative change during treatment for the external radiation plus 223Ra group (33 patients) was −65.6 % versus +9.3 % in the external beam radiation plus saline group (29 patients). This observation showed that the areas mostly affected by 223Ra were the regions with an elevated bone metabolism (Nilsson et al. 2007). In the external radiation plus 223Ra group, 15 of 31 patients had a prostate-specific antigen (PSA) decrease of more than 50 % from baseline compared to only 5 of 28 patients in the group receiving external radiation plus saline. The median time to PSA progression was 26 weeks in the 223Ra group and 8 weeks in the placebo group. A favourable adverse-event profile was confirmed with minimal bone marrow toxicity for patients who received 223Ra. Survival analyses from this phase 2 trial showed a significant OS benefit (Nilsson et al. 2007). The hazard ratio (HR) for survival, adjusted for baseline covariates, was 2.12 (P = .020, Cox regression).
A long-term clinical follow-up study from this phase 2 trial showed that more than twice as many patients receiving Alpharadin™ were alive (10 of 33) 2 years following start of treatment compared to those that received placebo (4 of 31). Analyses on 24-month OS and safety from the period 12–24 months after the first injection of study medication has recently been published (Nilsson et al. in press). Patients who received at least one dose of study medication had a median OS of 65 weeks in the radium-223 group versus 46 weeks in the placebo group (log-rank P = .056). The hazard ratio (HR) for OS, adjusted for baseline covariates, was 0.476 (95 % confidence interval [CI], 0.258–0.877; Cox regression P = .017). The most frequent cause of death for both arms was disease progression. There were no reports of treatment-related AEs or long-term haematological toxicity during the 12- to 24-month follow-up. Radium-223 had a highly favourable safety profile, with no evidence of second malignancies at 24-month follow-up (Nilsson et al. in press). The significant improvement in OS observed in patients receiving 223Ra versus placebo suggests a survival benefit.
BC 1–03. This clinical trial explored the dose–response relationship on pain palliation of 223Ra. A total of 100 patients with CRPC and painful bone metastases were randomized to a single intravenous dose of 5, 25, 50 or 100 kBq/kg. The primary endpoint was pain index (visual analogue scale or VAS and analgesic use) used to classify patients as responders or non-responders.
A significant dose response for pain index was seen at week 2 (P = .035) and with 40, 63, 56 and 71 % pain responders at week 8 (reduced pain and stable analgesic consumption) in the 5, 25, 50 and 100 kBq/kg groups, respectively (Nilsson et al. 2012). On the daily VAS, at week 8, pain decreased by a mean of −30, −31, −27 and −28 mm, respectively (P = .008, P = .0005, P = .002, and P < .0001) in these responders (post hoc analysis). A significant improvement in the brief pain inventory functional index was observed for all dose groups. Furthermore, a decrease in bone alkaline phosphatase in the higher dose groups was demonstrated (P = .0067). All doses were safe and well tolerated (Nilsson et al. 2012). The highly tolerable side-effect profile of 223Ra previously reported was confirmed.
BC 1–04. The aim of this phase 2 double-blind, dose-finding, multi-centre study was to prospectively evaluate the efficacy and safety of three different doses of 223Ra in patients with CRPC and bone metastases (Parker et al. 2013). A total of 122 patients were randomized to receive three injections with 6-week intervals, at doses of 25 kBq/kg (n = 41), 50 kBq/kg (n = 39), or 80 kBq/kg (n = 42). The study compared the proportion of patients in each dose group who had a confirmed decrease of ≥50 % in baseline PSA levels. Efficacy was evaluated using blood samples to measure PSA and other tumour markers, recorded skeletal-related events and pain assessments. Safety was evaluated using adverse events, physical examination and clinical laboratory tests.
A statistically significant dose–response relationship in confirmed ≥50 % PSA declines was found; for no patients (0 %) in the 25-kBq/kg dose group, two patients (6 %) in the 50-kBq/kg dose group and five patients (13 %) in the 80-kBq/kg dose group (P = .0297). A ≥50 % decrease in bone alkaline phosphatase levels was identified in 6 patients (16 %), 24 patients (67 %) and 25 patients (66 %) in the 25-, 50- and 80-kBq/kg dose groups, respectively (p < .0001). The most common treatment-related AEs (≥10 %) occurring up to week 24 across all dose groups were diarrhoea (21 %), nausea (16 %) and anaemia (14 %). No difference in incidence of haematological events was seen among dose groups (Parker et al. 2013). Radium-223 showed a dose-dependent effect on serum markers of CRPC activity, suggesting that control of bone disease with 223Ra may affect cancer-related outcomes.
6.3 Phase 3 Trial
Based on the favourable safety profile, improvements in disease-related biomarkers, pain palliation and a suggested OS benefit observed in the phase 2 studies mentioned above, the randomized, double-blind, multinational study (Parker et al. 2012; clinicaltrials.gov/ct2/show/NCT00699751) the so-called Alsympca-trial, was conducted in bone-metastatic CRPC-patients. The dosing regimen for this phase 3 trial was 50 kBq/kg every 4 weeks for six cycles. With a 2:1 randomization of the compound + best standard of care versus placebo + best standard of care, eligible patients were those regarded as unfit for or having failed chemotherapy with docetaxel. The primary endpoint was overall survival. At a pre-planned interim analysis 223Ra significantly improved survival versus placebo (14.0 months vs. 11.2 months; P = .00185). At the recently published cross-over analysis (n = 528 deaths) of all randomized patients the median OS benefit was 3.6 months; P = .00007 (Parker et al. 2012). In this study with 223Ra, analysis of OS was stratified by those receiving concurrent bisphosphonates or not. This 921 patient prospective trial indicated that the subset of patients treated with prior bisphosphonate had an HR for OS of 0.582 (95 % CI 0.397–0.8545-5) as compared to an HR of 0.752 (95 % CI 0.567–0.999). Direct comparison is not appropriate, but clearly there was a trend toward better efficacy in patients treated with a combination of 223Ra and bisphosphonates (Parker et al. 2012).
The mild and transient myelosuppression observed after 223Ra treatment was modest and with no signs of cumulative toxicity over the 6-monthly injections. This is different from that observed with the beta-emitting nuclides, where unpredictable and a long-lasting pancytopenia limit repeated treatments (Sartor et al. 2013). With 223Ra, the neutrophils decreased more than thrombocytes (Parker et al. 2012; Bruland et al. 2008b; Nilsson et al. 2007), whereas for beta-emitters, thrombocytopenia is commonly dose limiting.
7 Mechanisms of Tumour Targeting and Some Radiobiological Aspects
Due to short particle track length and potent cell-killing, an alpha-emitting BSR will, in contrast to the beta-emitters, deliver a much more energetic and localised radiation that produces densely ionising tracks and predominantly non-reparable double DNA-strand breaks (Bruland et al. 2006; Wheldon and O’Donoghue 1990; Henriksen et al. 2003; Li et al. 2004; Liepe 2009). With alpha-emitters, the endosteal bone surface received high radiation doses, whereas considerable fractions of the bone marrow were spared.