Radium‐223 Safety, Efficacy, and Concurrent Use with Abiraterone or Enzalutamide: First U.S. Experience from an Expanded Access Program

This article presents safety and exploratory efficacy findings from a U.S. expanded access program. Additionally, subgroup analyses evaluating overall survival by number of radium‐223 injections and safety and overall survival by prior or concomitant abiraterone or enzalutamide are reported.

Abstract

Background.

In the phase III ALSYMPCA trial, metastatic castration‐resistant prostate cancer (mCRPC) patients had few prior life‐prolonging therapies. Following ALSYMPCA, which demonstrated radium‐223 survival benefit, and before radium‐223 U.S. commercial availability, an expanded access program (EAP) providing early‐access radium‐223 allowed life‐prolonging therapies in current use.

Subjects, Materials, and Methods.

This phase II, open‐label, single‐arm, multicenter U.S. EAP (NCT01516762) enrolled patients with symptomatic mCRPC, ≥2 bone metastases, and no lung, liver, or brain metastases. Patients received radium‐223 55 kBq/kg intravenously every 4 weeks × 6. Primary outcomes were acute and long‐term safety. Additional analyses were done by number of radium‐223 injections, and prior or concomitant abiraterone or enzalutamide use.

Results.

Of 252 patients, 184 received radium‐223: 165/184 (90%) had Eastern Cooperative Oncology Group (ECOG) performance status 0–1; 183 (99%) had prior systemic anticancer therapy. Treatment‐related adverse events occurred in 93/184 (51%) patients during treatment and 11 (6%) during follow‐up. Median overall survival was 17 months, with 134/184 (73%) patients censored because of short follow‐up due to radium‐223 approval. In post hoc analyses, patients with ≥3 prior anticancer medications, baseline ECOG performance status ≥2, and lower baseline hemoglobin were less likely to receive 5–6 radium‐223 injections and unlikely to benefit from radium‐223. Radium‐223 was well tolerated regardless of concurrent or prior abiraterone or enzalutamide.

Conclusion.

Radium‐223 was well tolerated, with no new safety concerns; safety was maintained with abiraterone or enzalutamide. Patients with more advanced disease were less likely to benefit from radium‐223. Clinicians should consider baseline characteristics and therapy sequence for greatest clinical value.

Implications for Practice.

In this phase II U.S. expanded access program, radium‐223 was well tolerated, with a median overall survival of 17 months in metastatic castration‐resistant prostate cancer patients. In post hoc analyses, radium‐223 was safe regardless of concurrent abiraterone or enzalutamide, and median overall survival appeared longer when radium‐223 was used earlier in patients with less prior treatment. Patients with more advanced disease were less likely to benefit from radium‐223. Clinicians should consider baseline clinical characteristics and therapy sequence to provide the greatest clinical value to patients.

Introduction

In the 3 years since 2010, the number of life‐prolonging treatments for metastatic castration‐resistant prostate cancer (mCRPC) increased from one (docetaxel) to six, dramatically expanding treatment options and creating the potential to combine therapies for mCRPC. These options are docetaxel [1], sipuleucel‐T [2], cabazitaxel [3], abiraterone acetate (abiraterone) [4], enzalutamide [5], and radium‐223 dichloride (radium‐223), a targeted alpha therapy [6]; each improved overall survival (OS) versus placebo or standard of care in randomized controlled trials. In the phase III ALSYMPCA trial (NCT00699751), radium‐223 versus placebo significantly prolonged OS (hazard ratio [HR] = 0.70; 95% confidence interval [CI] 0.58–0.83; p < .001) [6], delayed time to first symptomatic skeletal event (SSE; HR = 0.66; 95% CI 0.52–0.83; p < .001) [7], resulted in meaningful improvement in quality of life [8], and was well tolerated, with a low myelosuppression incidence [9].

Following the ALSYMPCA demonstration of OS benefit with radium‐223 and before the agent's commercial availability in the U.S., a phase II, prospective, interventional, open‐label, multicenter expanded access program (EAP) provided early access to radium‐223 for mCRPC, and monitored its acute and long‐term safety outside a registration trial in a context reflecting current routine clinical use. The U.S. EAP was part of a larger worldwide research effort including the international EAP, which treated 696 patients from 14 countries [10].

The ALSYMPCA trial started in 2008 [6], and eligible patients were pretreated with few life‐prolonging therapies. In contrast, U.S. EAP patients (trial began in 2012) had access to agents with a demonstrated survival advantage in mCRPC; these agents were permitted before enrollment. Additionally, abiraterone and enzalutamide were allowed as concomitant therapies during radium‐223 treatment.

Here we present safety and exploratory efficacy findings from U.S. EAP. Additionally, we present subgroup analyses evaluating OS by number of radium‐223 injections, and safety and OS by prior or concomitant abiraterone or enzalutamide.

Subjects, Materials, and Methods

Patients

Eligibility criteria were similar to those of ALSYMPCA [6]. Patients had symptomatic, progressive, bone‐predominant mCRPC with ≥2 bone metastases on imaging with no lung, liver, or brain metastases. Lymph node metastases were allowed. Symptomatic was defined as regular (not occasional) use of any analgesic medication (including acetaminophen) for cancer‐related bone pain (≥level 1; World Health Organization ladder for cancer pain) or treatment with external beam radiation therapy for bone pain within 12 weeks before treatment. Progressive was defined as the appearance of new bone lesions or two subsequent increases in serum prostate‐specific antigen (PSA) over the previous reference value. Any bone‐imaging techniques were permitted to evaluate disease progression in bone, per institutional standard of care. Additional eligibility criteria included age ≥18 years; no intention to use cytotoxic chemotherapy within the next 6 months; life expectancy ≥6 months; Eastern Cooperative Oncology Group (ECOG) performance status 0–2; and adequate hematologic, liver, and renal function.

Patients were excluded if they had an investigational drug within the previous 4 weeks, were eligible for a first docetaxel course, had received cytotoxic chemotherapy within 4 weeks before screening, or failed to recover from adverse events (AEs) associated with prior chemotherapy (ongoing neuropathy was permitted). Additional exclusion criteria included prior hemibody external radiotherapy, prior systemic therapy with radionuclides, presence of visceral or brain metastases, lymphadenopathy >6 cm in short‐axis diameter, or any‐size pelvic lymphadenopathy thought to contribute to concurrent hydronephrosis, imminent spinal cord compression, or fecal incontinence. All patients provided written informed consent.

Study Design and Treatment

The U.S. EAP was a prospective, interventional, open‐label, single‐arm, multicenter phase II EAP conducted at 26 U.S. centers (supplemental online Fig. 1). Patients received radium‐223 (50 kBq/kg; 55 kBq/kg intravenously after U.S. National Institutes of Standards and Technology update [11]) every 4 weeks for six injections. Radium‐223 was added to best standard of care as provided per institutional guidelines at each study center. Patients who did not have bilateral orchiectomy were to receive luteinizing hormone–releasing hormone agonists or polyestradiol phosphate throughout the study. Cytotoxic chemotherapy, other systemic radioisotopes, hemibody external radiotherapy, or other investigational drug was not permitted during the treatment period, but was allowed during follow‐up when administered ≥4 weeks after last study‐drug injection.

The treatment period was the time from first radium‐223 dose to 30 days after the last dose. Visits occurred each cycle (i.e., every 4 weeks ±7 days) for treatment and evaluation of ECOG performance status, clinical laboratory tests, SSEs, and AEs. Laboratory assessments were performed locally and evaluated within 72 hours before study‐drug administration. All patients who completed or discontinued treatment were eligible to participate in the follow‐up, every 6 months until death or study termination upon radium‐223 commercial availability.

The sponsor provided radium‐223, manufactured by the Institute for Energy Technology, Isotope Laboratories, Kjeller, Norway. Study‐drug dose level adjustments were not permitted, and every effort was made to keep the treatment interval at 4 weeks ±7 days. Treatment delay due to an AE could be up to 4 weeks. Patients were withdrawn from the study if they had a dose delay of >4 weeks (maximum 8 weeks between radium‐223 injections); if they had disease progression, any intolerable AE, or grade 3 or 4 anemia, neutropenia, or thrombocytopenia lasting >14 days; or by patient or investigator request. Withdrawn patients were not replaced; however, follow‐up data collection continued.

The study follow‐up was short due to radium‐223 U.S. Food and Drug Administration (FDA) approval on May 15, 2013. Patients who were still undergoing treatment at the time of approval were allowed to finish treatment, but were only followed for 30 days after last radium‐223 dose. The study terminated within 8–12 weeks of FDA approval.

Review boards at all participating centers approved the study, which was conducted in accordance with the Declaration of Helsinki and Good Clinical Practice guidelines of the International Conference on Harmonization and registered at ClinicalTrials.gov, number NCT01516762.

Outcomes

Primary U.S. EAP outcomes were acute and long‐term safety of radium‐223. Safety variables during the treatment period included ECOG performance status, SSEs, grades 3–5 treatment‐emergent adverse events (TEAEs), any‐grade treatment‐related AEs, any‐grade TEAEs leading to study‐drug discontinuation, and serious AEs (SAEs; definitions in supplemental online Table 1). Safety variables during follow‐up were SSEs, treatment‐related AEs and SAEs, and additional primary malignancies. Exploratory efficacy variables included OS, time to first SSE, time to disease progression, time to total alkaline phosphatase (tALP) or PSA progression, tALP or PSA changes, and tALP normalization (definitions in supplemental online Table 1).

Assessments

During the treatment period, TEAEs, treatment‐related AEs, TEAEs leading to drug discontinuation, and SAEs were recorded. All cases of leukemia, myelodysplastic syndrome, aplastic anemia, primary bone cancer, or other second primary malignancy were reported as SAEs regardless of the investigator's assessment of relationship to study drug. During follow‐up (i.e., ≥30 days after last treatment), AEs and SAEs were reported only if considered treatment related by the investigator. AEs were graded per the Common Terminology Criteria for Adverse Events version 4.03. Exploratory efficacy assessments included survival status, clinically evaluated SSEs, and tALP and PSA concentrations.

Statistical Analysis

No statistical assumptions were made regarding study sample size. Safety and exploratory efficacy analyses were performed in the safety population, including all patients who received ≥1 radium‐223 injection. Descriptive statistics were used to summarize patient demographics and safety. Time‐to‐event data were summarized using Kaplan‐Meier methodology. Statistical analyses used SAS versions 9.2 and 9.3 (SAS Institute Inc., Cary, NC).

Our exploratory endpoints assessing OS included a post hoc subgroup analysis of number of injections, with a stepwise logistic regression analysis of baseline covariates predicting number of injections received. Following the stepwise approach, the significance limit for entry into the model was set at 0.20; the significance limit to stay in the model after previous entry was set at 0.35. After assessing effects on OS, we performed a logistic regression of baseline covariates to determine their association with receiving 1–4 versus 5–6 radium‐223 injections. We examined the following baseline parameters: ≥3 prior anticancer medications (yes/no), ECOG ≥2 (yes/no), albumin (g/dL), log hemoglobin (g/dL), log PSA (log µg/L), and log ALP (log U/L). To convert albumin and hemoglobin to g/L, multiply by 10.

In post hoc analyses of prior or concurrent abiraterone or enzalutamide, prior treatment was defined as abiraterone or enzalutamide taken and stopped before first radium‐223 dose; concurrent was defined as treatment started and received with radium‐223 or received before and with radium‐223 up to 30 days after last radium‐223 dose.

For all post hoc analyses, baseline characteristics and disposition for patient subgroups were summarized using descriptive statistics; OS analysis used Kaplan‐Meier methodology.

Results

Patients

From April 11, 2012, through July 10, 2013, 252 patients were enrolled from 26 U.S. centers; 184 received at least one radium‐223 dose and were included in the safety population (supplemental online Fig. 2). Of 184 patients, 135 (73%) were aged ≥65 years, 165 (90%) had ECOG performance status 0–1, and 183 (99%) had prior systemic anticancer therapy, including 26 (14%) with prior abiraterone alone, 31 (17%) with prior abiraterone plus docetaxel, and 29 (16%) with prior abiraterone plus docetaxel plus enzalutamide (Table 1).

Table 1. Patient demographics and baseline clinical characteristics.

^aData are from the safety population (n = 184) unless otherwise stated; percentages may not sum to 100 due to rounding.

^bAssessed by Brief Pain Inventory—short form.

^cPatients may have received more than one type of procedure or medication.

^dDrugs shown may have been given with other drugs that are not specified in the table; 37 (20%) patients did not receive any of the four anticancer therapies shown (i.e., abiraterone, cabazitaxel, docetaxel, or enzalutamide).

^eThe docetaxel record of missing imputed start and stop dates for one patient was not included in the calculation of duration.

Abbreviations: ALP, alkaline phosphatase; ECOG PS, Eastern Cooperative Oncology Group performance status; mo, months; PSA, prostate‐specific antigen; ULN, upper limit of normal.

The median number of study‐drug injections received was 5 (range, 1–6), and 81 of 184 (44%) patients received all 6 radium‐223 injections. Among 103 patients who discontinued treatment early, the primary reason was disease progression in 32 (31%) patients (supplemental online Fig.2).

Safety

The majority of patients within each treatment cycle had no change in ECOG performance status during the treatment period (supplemental online Fig. 3). More than 1 TEAE (any grade) occurred in 133 of 184 (72%) patients, and 75 (41%) patients had grade 3–5 TEAEs during treatment (Table 2).

Table 2. TEAEs during the treatment period.

^aData are n (%) of patients with an adverse event.

Abbreviation: TEAE, treatment‐emergent adverse events.

Treatment‐related AEs (all grades) occurred in 93 of 184 (51%) patients during the treatment period and 11 (6%) during follow‐up; the most common during treatment were anemia (16%), fatigue (15%), diarrhea (13%), nausea (10%), and decreased platelet count (8%; Table 3). TEAEs (any grade) leading to treatment discontinuation occurred in 30 of 184 (16%) patients during treatment (supplemental online Table 2). SAEs occurred in 45 of 184 (25%) patients during treatment and 9 (5%) during follow‐up (supplemental online Table 3). AEs of a hemorrhagic nature that occurred during treatment were upper gastrointestinal hemorrhage (grade 4, considered treatment‐related), rectal hemorrhage (grade 3, associated with thrombocytopenia and considered treatment‐related [led to study‐drug discontinuation]), subdural hematoma (grade 3, associated with thrombocytopenia and considered treatment‐related), hematuria (grade 2), and epistaxis (grade 1, associated with thrombocytopenia), each in one patient and none with a fatal outcome. Additional primary malignancies occurred in 3 of 184 (2%) patients: basal cell carcinoma, plasma cell myeloma, and bronchial carcinoma, each in 1 patient and none attributed to radium‐223.

Table 3. Treatment‐related AEs.

^aData are n (%) of patients with an AE.

^bIncludes AEs related to study drug that started more than 30 days after the last injection of radium‐223.

^cOccurring in ≥5% of patients during treatment period.

Abbreviation: AE, adverse event.

Exploratory Efficacy

Median OS was 17 months (95% CI, 11 months–not estimable), with 50 deaths and a high percentage of patients censored (134/184 [73%]) because of short follow‐up due to radium‐223 approval (Fig. 1A). The most common cause of death was progressive disease in 35 of 50 (70%) patients. Additional prespecified exploratory efficacy variables are shown in Table 4.

Figure 1.

Kaplan‐Meier estimates of overall survival for the total U.S. expanded access program population and patient subgroups. (A): Safety population. (B): Number of radium‐223 injections. (C): Prior abi or enza. (D): Abi and enza naïve^a. (E): Concurrent abi or enza as first‐line therapy^b. (F): Concurrent abi or enza as second‐line therapy^c. ^aNaïve was defined as patients with no prior use of abiraterone and enzalutamide. ^bConcurrent abiraterone with no prior enzalutamide or concurrent enzalutamide with no prior abiraterone. ^cConcurrent abiraterone with prior enzalutamide or concurrent enzalutamide with prior abiraterone.Abbreviations: abi, abiraterone; enza, enzalutamide; mo, months; NE, value cannot be estimated due to censored data.

Table 4. Exploratory efficacy variables.

^aIncludes all treatment‐emergent SSEs. Multiple SSE components could have occurred at any time or on the same date, and a patient may have been counted in more than one category.

^bDisease progression was based on routine clinical practice evaluation.

^cConfirmed by a second value approximately 4 or more weeks later.

^dNormalization defined as return of the tALP value to within normal range at 12 weeks in two consecutive measurements (at least 2 weeks apart) after the start of treatment in patients with tALP above the ULN at baseline.

Abbreviations: CI, confidence interval; EBRT, external beam radiation therapy; mo, months; NE, value cannot be estimated due to censored data; PSA, prostate‐specific antigen; SSE, symptomatic skeletal event; tALP, total alkaline phosphatase; ULN, upper limit of normal.

OS by Number of Radium‐223 Injections

In qualitative data reviews, the number of radium‐223 injections administered were notably distinct. Among 184 patients, 85 (46%) received 1–4 radium‐223 injections and 99 (54%) received 5–6 radium‐223 injections (18 of 184 [10%] received 5 injections; 81 of 184 [44%] received all 6 injections). Patients with 5–6 versus 1–4 injections had less advanced disease, evidenced by a lower percentage of patients with prior docetaxel (51% vs. 71%), severe baseline pain (13% vs. 26%), ECOG performance status of 2 (4% vs. 14%), lower baseline levels of PSA (median 121 µg/L vs. 186 µg/L) and tALP (median 132 U/L vs. 170 U/L), and a longer median time since prostate cancer diagnosis (median 7.1 years vs. 5.9 years; Table 5). Premature treatment discontinuation was primarily due to disease progression for patients with 1–4 injections (30 of 85 patients [35%]) and AEs for patients with 5 injections (8 of 18 patients [44%]; supplemental online Table 4).

Table 5. Patient demographics and baseline clinical characteristics, by number of study drug injections.

Abbreviations: ALP, alkaline phosphatase; ECOG PS, Eastern Cooperative Oncology Group performance status; mo, months; PSA, prostate‐specific antigen.

Median OS for patients receiving 1–4 versus 5–6 radium‐223 injections was 7.5 months (95% CI, 6.5–8.2 months) versus not reached (Fig. 1B). Baseline covariates predicting receipt of 1–4 versus 5–6 radium‐223 injections were ≥3 prior anticancer medications, ECOG performance status ≥2, and lower hemoglobin levels (Table 6).

Table 6. Logistic regression analysis of baseline covariates predictive of patients receiving 1–4 versus 5–6 injections of radium‐223.

^aHemoglobin was assessed as a continuous variable.

Abbrevation: ECOG PS, Eastern Cooperative Oncology Group performance status.

Prior Abiraterone or Enzalutamide

Among 184 patients, 83 (45%) had received prior abiraterone or enzalutamide, 48 (26%) had prior abiraterone and enzalutamide, and 53 (29%) were abiraterone‐ or enzalutamide‐naïve. Prior docetaxel had also been administered to 54 of 83 (65%) patients with prior abiraterone or enzalutamide, and 16 of 53 (30%) who were abiraterone‐ or enzalutamide‐naïve (supplemental online Table 5). Patients with prior abiraterone or enzalutamide had a median of 5 (range, 1–6) radium‐223 injections, with 37 of 83 (45%) receiving all 6 injections. Abiraterone‐ or enzalutamide‐naïve patients had a median of 6 (range, 2–6) radium‐223 injections, with 30 of 53 (57%) receiving all 6 injections.

Radium‐223 was well tolerated regardless of prior abiraterone or enzalutamide (supplemental online Table 6). Kaplan‐Meier OS plots by prior abiraterone or enzalutamide are shown in Figure 1C and 1D.

Concurrent Abiraterone or Enzalutamide

Concurrent abiraterone was administered to 25 of 184 (14%) patients, and concurrent enzalutamide to 15 of 184 (8%) patients. Among patients with concurrent abiraterone, 8 of 25 (32%) had prior enzalutamide and 11 of 25 (44%) had prior docetaxel. Among patients with concurrent enzalutamide, 14 of 15 (93%) had prior abiraterone and 11 of 15 (73%) had prior docetaxel. Baseline characteristics of patients who received concurrent abiraterone or enzalutamide were generally similar, except for higher baseline PSA levels (322 µg/L vs. 94 µg/L, respectively) and prior docetaxel use (73% vs. 44%, respectively) in patients with concurrent enzalutamide versus concurrent abiraterone (supplemental online Table 7).

Radium‐223 concurrently administered with abiraterone or enzalutamide was well tolerated (supplemental online Table 8). The most frequent grade 3–4 TEAEs were anemia (abiraterone 16%, enzalutamide 13%), thrombocytopenia (abiraterone 4%, enzalutamide 0%), and back pain (abiraterone 0%, enzalutamide 13%). Kaplan‐Meier OS plots by concurrent abiraterone or enzalutamide use are shown in Figure 1E and 1F.

Discussion

In this early access, open‐label, phase II U.S. EAP study, radium‐223 was well tolerated, with no new safety concerns in mCRPC patients. Radium‐223 showed a positive effect on the exploratory efficacy parameters, with a median OS of 17 months. These results are similar to ALSYMPCA [6] and international EAP [10] safety and efficacy findings, providing both valuable data from patients representative of those currently treated in U.S. clinical practice and insight into radium‐223 use within the evolving mCRPC treatment paradigm.

This study focused on safety in a U.S. population, whereas phase III ALSYMPCA and phase IIIb international EAP had safety and OS as primary objectives in patients mainly from outside the U.S. U.S. EAP, unlike ALSYMPCA, permitted prior exposure to agents with a demonstrated survival advantage in mCRPC; 147 of 184 (80%) U.S. EAP patients had prior abiraterone, enzalutamide, cabazitaxel, or docetaxel, indicative of a population with more advanced disease than in ALSYMPCA. Because these agents were permitted before U.S. EAP enrollment, and abiraterone and enzalutamide were allowed concomitantly, this study generates important safety and efficacy data regarding sequencing and combination with radium‐223.

Post hoc findings revealed that radium‐223 was well tolerated regardless of concurrent or prior abiraterone or enzalutamide, suggesting the possibility of combining targeted alpha therapy with existing mCRPC agents. OS appeared better when radium‐223 was used earlier in patients with less prior treatment, regardless of whether they received radium‐223 in sequence or combination with abiraterone or enzalutamide. These compelling preliminary findings are hypothesis‐generating, requiring future studies to change practice; however, when considered in light of recent results from the STAMPEDE and LATITUDE trials showing an OS benefit to adding abiraterone to androgen deprivation therapy in patients with high‐risk early‐stage (e.g., hormone‐naïve) metastatic prostate cancer, it raises the hypothesis that using radium‐223 in combination with abiraterone in the upfront setting may be a reasonable and testable hypothesis for treating patients with bone‐metastatic disease [12], [13]. Given the nature of this EAP, study limitations include no comparator group, small patient numbers, few death events in the OS analysis, potential patient‐selection bias, and short follow‐up. Additionally, radium‐223 may not have been initiated with abiraterone or enzalutamide, but rather started when other drugs began to fail or PSA began to rise before radiographic progression. These timing nuances and issues of disease progression (serologic or radiographic) cannot be dissected within this database. Currently, evidence is limited that we have observed anything other than a marker of earlier disease and less extensive prior therapy. Efficacy and safety of radium‐223 combined with abiraterone or enzalutamide are being investigated in one open‐label phase II trial (NCT02097303), one randomized phase II trial (NCT02034552), and two phase III trials (NCT02043678 and NCT02194842).

Evidence for better survival outcomes with earlier radium‐223 was also demonstrated in post hoc analyses of number of injections received in the U.S. EAP. Median OS appeared to be prolonged with 5–6 versus 1–4 radium‐223 injections (median not reached vs. 7.5 months, respectively). In similar analyses of ALSYMPCA and international EAP patients, 5–6 versus 1–4 study‐drug injections was also associated with prolonged OS [14], [15]. In all three studies, there appeared to be an association between number of radium‐223 injections and OS, with 5–6 injections associated with prolonged survival; however, because number of injections and OS are postbaseline factors, these results could be confounded by patient baseline characteristics. Additionally, treatment exposure may result from better baseline characteristics and more opportunity to receive drug, prompting a logistic regression analysis of baseline covariates to understand factors contributing to receiving 5–6 radium‐223 injections. We found that patients with more advanced disease (i.e., ≥3 prior anticancer medications, baseline ECOG performance status ≥2, and lower baseline hemoglobin) were unlikely to benefit from radium‐223. Radium‐223 treatment should be considered earlier in patients with less advanced disease and fewer lines of prior therapy to ensure receiving 5–6 injections. Although these post hoc results are hypothesis‐generating and require validation in a prospective randomized trial, it is notable that baseline parameters and number of prior therapies may influence a patient's ability to receive 5–6 radium‐223 injections.

Conclusion

Radium‐223 was well tolerated, with no new safety concerns, in this open‐label phase II study providing early access to radium‐223 for mCRPC patients. In post hoc analyses, the safety profile was maintained when radium‐223 was combined with abiraterone or enzalutamide. Patients with more advanced disease were less likely to benefit from radium‐223. It is unknown whether the association between number of injections and prolonged OS is due to patient selection or use in less advanced patients; however, clinicians should consider baseline clinical characteristics and therapy sequence to provide the greatest clinical value to patients.

See http://www.TheOncologist.com for supplemental material available online.

Author Contributions

Conception/design: Oliver Sartor, Nicholas J. Vogelzang, Matthew R. Smith, Jorge A. Garcia, Oana Petrenciuc, Neal D. Shore, Michael J. Morris

Provision of study materials or patients: Oliver Sartor, Nicholas J. Vogelzang, Christopher Sweeney, Daniel C. Fernandez, Fabio Almeida, Andrei Iagaru, Alan Brown Jr., Matthew R. Smith, Manish Agrawal, Adam P. Dicker, Jose Lutzky, Yu‐Ning Wong, Neal D. Shore, Michael J. Morris

Collection and/or assembly of data: Oliver Sartor, Nicholas J. Vogelzang, Christopher Sweeney, Daniel C. Fernandez, Yu‐Ning Wong

Data analysis and interpretation: Oliver Sartor, Nicholas J. Vogelzang, Christopher Sweeney, Matthew R. Smith, Manish Agrawal, Adam P. Dicker, Jose Lutzky, Yu‐Ning Wong, Oana Petrenciuc, Jeremy Gratt, Neal D. Shore, Michael J. Morris

Manuscript writing: Oliver Sartor, Nicholas J. Vogelzang, Christopher Sweeney, Daniel C. Fernandez, Alan Brown Jr., Matthew R. Smith, Manish Agrawal, Jorge A. Garcia, Jose Lutzky, Yu‐Ning Wong, Neal D. Shore, Michael J. Morris

Final approval of manuscript: Oliver Sartor, Nicholas J. Vogelzang, Christopher Sweeney, Daniel C. Fernandez, Fabio Almeida, Andrei Iagaru, Alan Brown Jr., Matthew R. Smith, Manish Agrawal, Adam P. Dicker, Jorge A. Garcia, Jose Lutzky, Yu‐Ning Wong, Oana Petrenciuc, Jeremy Gratt, Neal D. Shore, Michael J. Morris

Disclosures

Oliver Sartor: Bayer, Bellicum, Bristol‐Myers Squibb, Celgene, Dendreon, Johnson & Johnson, Medivation, OncoGeneX, Sanofi‐Aventis, Tokai (C/A), Bayer, Dendreon, Johnson & Johnson, Sanofi‐Aventis, Endocyte, Innocrin (RF); Nicholas J. Vogelzang: Algeta/Bayer (RF), Bayer (H); Christopher Sweeney: Sanofi, Janssen, Astellas, Bayer (RF, H), Pfizer (H); Andre Iagaru: Bayer HealthCare, GE Healthcare, Piramal Imaging GmbH (RF); Alan Brown Jr.: Bayer HealthCare (H); Matthew R. Smith: Bayer (C/A); Adam P. Dicker: Bayer (H); Jorge A. Garcia: Bayer (RF, H), Bayer, Medivation/Astellas (C/A), Johnson & Johnson, Medivation/Astellas (RF, institution); Jose Lutzky: Bayer (RF); Yu‐Ning Wong: Janssen Scientific Affairs (Johnson & Johnson) (E); Fox Chase Cancer Center, Temple University Health System (E, at time of study), Bayer (RF, institution [FCCC]), Pfizer (RF), PhRMA Foundation (C/A); Oana Petrenciuc: Bayer HealthCare Pharmaceuticals (E); Jeremy Gratt: Bayer HealthCare (H); Neal D. Shore: Amgen, Astellas, AbbVie, Bayer, Dendreon, Ferring, Janssen, Medivation, Sanofi Genzyme, Tolmar, Myovant (C/A, RF); Michael J. Morris: Bayer, Endocyte, Progenics (RF), Bayer, Astellas, Endocyte (SAB, noncompensated), Progenics, Millennium (SAB, compensated). The other authors indicated no financial relationships.

(C/A) Consulting/advisory relationship; (RF) Research funding; (E) Employment; (ET) Expert testimony; (H) Honoraria received; (OI) Ownership interests; (IP) Intellectual property rights/inventor/patent holder; (SAB) Scientific advisory board