Real-world effectiveness and safety of radium-223 therapy in castration-resistant prostate cancer with bone metastases

Abstract

The effectiveness and safety of radium-223 (Ra-223) in patients with metastatic castration-resistant prostate cancer (mCRPC) and bone metastasis in real-world clinical practice in China remain to be fully examined. This single-center, retrospective, real-world study included Chinese patients with mCRPC treated with Ra-223 between January 2021 and August 2023 in Fudan University Shanghai Cancer Center (Shanghai, China). This study examined the patient characteristics, overall survival (OS), 5-point Visual Analog Scale for pain and safety, as well as factors potentially influencing treatment effectiveness. Ninety-three patients were included. Among them, 25 (26.9%) patients were administered all six injections of Ra-223, and 41 (44.1%) patients received more than 3 injections. After a median follow-up of 19.8 months, median OS was 14.1 (95% confidence interval [CI]: 12.0–16.2) months. Multivariable analysis showed a total number of Ra-223 injections >3 was an independent protective factor of OS (hazard ratio [HR] = 0.443, 95% CI: 0.234–0.837; P = 0.012) after adjusting for age, combination therapy, and previous chemotherapy. For those administered all six injections of Ra-223, no patients had a pain score of 4 and only 4.2% of patients showed a pain score of 3. In total, 62.5% of patients had almost no pain (0 point). Median alkaline phosphatase showed a downward trend during the six-injection treatment period. The most common adverse event was anemia (30.1%), followed by thrombocytopenia (15.1%), nausea (9.7%), weakness (8.6%), and decreased white blood cells (7.5%). This real-world study strongly suggests that Ra-223 has good effectiveness and tolerability in Chinese patients with mCRPC and bone metastases.

INTRODUCTION

It is estimated that the annual incidence of metastatic castration-resistant prostate cancer (mCRPC) will rise from 1.4 million in 2020 to 2.9 million by 2040.1 mCRPC is the lethal form of prostate cancer progression.2 Bones are the most common sites of metastasis (84%) in prostate cancer besides lymph nodes. Bone metastases can cause bone pain, pathological fractures, and spinal cord compression, further worsening the prognosis.3

Radium-223 (Ra-223) is a short-range-alpha-particle emitter that, as a calcium mimetic, accumulates in bone metastases with high bone turnover.4 Alpha radiation could induce the formation of targeted double-strand DNA breaks and exert other cytotoxic effects in tumors, while its systemic toxicity and hematologic complications are limited by the relatively short diffusion range of the radiation.5,6 The National Comprehensive Cancer Network (NCCN) guidelines recommend using Ra-223 for mCRPC patients with symptomatic bone metastases based on substantial clinical benefits. The phase III Alpharadin in Symptomatic Prostate Cancer Patients (ALSYMPCA) trial showed that compared with placebo, Ra-223 improved overall survival (OS) in patients with mCRPC (14.9 months vs 11.3 months) as well as the quality of life (QOL), with good tolerability.7,8,9 Furthermore, several studies have reported improvements in QOL in patients with mCRPC treated with Ra-223.10,11,12 Subsequently, Ra-223 was approved by the U.S. Food and Drug Administration (FDA)13 and National Medical Products Administration of China.

To date, treatment strategies for Ra-223 are still under exploration, including in combination with other drugs such as androgen receptor pathway inhibitors (ARPIs),14 olaparib,15 and atezolizumab.16 Furthermore, several factors affecting the real-world effectiveness of Ra-223 cannot be captured by randomized control trials (RCTs). It is well-documented that Asian patients with prostate cancer feature genomic abnormalities that are distinct from those cases from Western countries,17 which is reflected by relatively better prognosis across different treatment regimens.18,19 However, the evaluation of Ra-223 benefits in the Asian population is lacking. A study has shown that the probability of Ra-223 being cost-effective for patients with bone metastases is sharply low in China, as a developing country.20 Indeed, many patients are unable to complete the standard six-injection Ra-223 treatment,21,22 possibly due to both physical reasons and cost issues.

To examine the effectiveness and safety of Ra-223 under real-world circumstances, we retrospectively analyzed 93 patients administered Ra-223 in Fudan University Shanghai Cancer Center (Shanghai, China) from January 2021 to August 2023. Given that Ra-223 was just authorized by the National Medical Products Administration (NMPA) in August 2020, this study provides timely real-world evaluation of Ra-223 in the Chinese mainland. Our results could provide evidence for managing patients with mCRPC.

PATIENTS AND METHODS

Patients and study design

This single-center, retrospective, real-world study in China included patients with mCRPC administered Ra-223 in Fudan University Shanghai Cancer Center between January 2021 and August 2023. Patient characteristics, OS, pain, and safety were evaluated, as well as potential factors affecting effectiveness.

Inclusion criteria were as follows: (1) CRPC diagnosis, based on the European Association of Urology guidelines (i.e., serum testosterone <50 ng dl⁻¹ with medical or surgical castration);23 (2) bone metastasis diagnosis, based on bone scintigraphy; (3) at least one injection of Ra-223, either as monotherapy or combination therapy; and (4) complete data determined by the investigators. All patients who received Ra-223 treatment in other hospitals before visiting the study hospital were excluded.

The study was approved by the Ethics Committee of Fudan University Shanghai Cancer Center (Approval No. 050432-4-2307E). The committee approved the publication of the anonymized data included in this article. The requirement for individual consent was waived by the ethics committee due to the retrospective nature of the study.

Treatment pattern, assessment of clinical outcomes, and follow-up

During the study period, a typical Ra-223 treatment was an intravenous injection of 55 kBq kg⁻¹ of Ra-223, on day 1 of every 4-week cycle, for up to six cycles. Each case was evaluated and discussed by a multidisciplinary team before reaching a decision on whether to administer ARPIs and chemotherapy. The actual number of Ra-223 injections administered, the treatment line of Ra-223 use, and data on the combination therapy were recorded to evaluate treatment patterns. Patients administered bisphosphonate or anti-receptor activator of nuclear factor-kappa B ligand (RANKL) monoclonal antibody were considered as having received bone-protecting agents.

The clinical outcomes were OS, pain relief, changes in biomarkers, and safety. OS was the time from the first Ra-223 injection to death from any cause. The vital status at follow-up and date of death, if applicable, were determined using routine outpatient/inpatient visit records and telephone follow-up. Pain records were extracted from the patient charts. In clinical practice, pain assessment was part of routine examinations and was conducted by the attending physician. In patients taking pain-control medications, each pain assessment was performed before the scheduled dosing. Pain scores were recorded using a 5-point Visual Analog Scale, with 0 representing no pain and 4 representing the most severe pain imaginable. The biomarker alkaline phosphatase (ALP) was routinely measured during the study period at each visit. Safety was evaluated based on the adverse events (AEs) recorded in patient charts and laboratory tests.

Statistical analyses

Missing data were not imputed in this study. Categorical variables were presented as n (%) and compared by the Chi-square test or the Fisher’s exact test. The distribution of continuous variables was tested by the Shapiro-Wilk test. Normally distributed data were presented as mean ± standard deviation (s.d.), whereas non-normally distributed data were presented as median (interquartile range [IQR]). Continuous data were compared by the two-independent-sample rank-sum test. Survival was analyzed by the Kaplan–Meier method, and survival curves were compared by the log-rank test. Cox regression analysis was used for univariable and multivariable analyses. Subgroup analyses were conducted based on combination therapy, treatment line, previous chemotherapy, and the number of Ra-223 injections. Two-sided P < 0.05 was considered statistically significant. Statistical analyses were conducted by R (version 3.5.1; R Foundation for Statistical Computing, Vienna, Austria).

RESULTS

Patient characteristics and treatments

Ninety-three patients were included. Median patients’ age was 70 (IQR: 64–76) years, and 21 (22.6%) cases had an Eastern Cooperative Oncology Group (ECOG) performance status (PS) of 3. Forty-five and 48 patients were at treatment lines ≤3 and >3, respectively. Only 25 (26.9%) patients completed the six-injection regimen, and 41 (44.1%) patients received >3 injections. Fifty-two patients were administered ≤3 injections totally due to early treatment discontinuation, predominantly due to disease progression (23.1%) and death during treatment (21.2%) as shown in Supplementary Table 1. Ra-223 monotherapy was administered in 60 (64.5%) patients. A bone-protecting agent was used in 50 (53.8%) patients (Table 1).

Supplementary Table 1.

Summary of causes of early discontinuation of treatment (patients who received ≤3 injections)

Cause of discontinuation	n (%)
Disease progression	12 (23.1)
Death during treatment	11 (21.2)
Perceived lack of efficacy	8 (15.4)
Intolerable hematologic toxicity	8 (15.4)
Financial constraints	6 (11.5)
Pain exacerbation	4 (7.7)
Intolerable gastrointestinal toxicity	3 (5.8)

Table 1.

Characteristics of the patients

Characteristic	Value (total patients: 93)
Age (year), median (IQR)	70.0 (64.0–76.0)
Age group, n (%)
<70 years	40 (43.0)
≥70 years	53 (57.0)
ECOG, n (%)
0	8 (8.6)
1	47 (50.5)
2	17 (18.3)
3	21 (22.6)
PSA (ng ml−1), median (IQR)	47.7 (7.8–134.0)
ALP (IU l−1), median (IQR)	146.0 (101.6–227.5)
Gleason score, n (%)
7	14 (15.1)
8	16 (17.2)
9	50 (53.8)
10	13 (14.0)
Pain grading, n (%)
0	13 (14.0)
1	24 (25.8)
2	26 (28.0)
3	29 (31.2)
4	1 (1.1)
Treatment line grouping, n (%)
Treatment line ≤3	45 (48.4)
Treatment line >3	48 (51.6)
Previous chemotherapy, n (%)
No	42 (45.2)
Yes	51 (54.8)
Total number of Ra-223 injections, n (%)
1	26 (28.0)
2	14 (15.1)
3	12 (12.9)
4	12 (12.9)
5	4 (4.3)
6	25 (26.9)
Total number of Ra-223 injections, n (%)
≤3	52 (55.9)
>3	41 (44.1)
Combination therapy, n (%)
No	60 (64.5)
Yes	33 (35.5)
Combination therapeutic drug, n (%)
ARi	21 (22.6)
Abiraterone	8 (8.6)
Docetaxel	3 (3.2)
PARPi	1 (1.1)
Using bone-protecting agents, n (%)
No	43 (46.2)
Yes	50 (53.8)

ECOG: Eastern Cooperative Oncology Group; PSA: prostate-specific antigen; ALP: alkaline phosphatase; ARi: androgen receptor inhibitor; PARPi: poly(ADP-ribose) polymerase inhibitor; IQR: interquartile range; Ra-223: radium 223

Overall survival and prognostic factors

The median follow-up was 19.8 months. The median OS of all patients was 14.1 (95% confidence interval [CI]: 12.0–16.2) months. The 12- and 24-month survival rates were 59.3% (95% CI: 48.9%–71.8%) and 23.0% (95% CI: 12.2%–43.6%), respectively (Figure 1a and Supplementary Table 2).

Figure 1.

Overall survival (OS) analysis in patients treated with Ra-223. (a) Kaplan–Meier curve analysis showed a median OS of 14.1 months in all patients. (b) The impact of previous chemotherapy on OS was demonstrated, with patients without prior chemotherapy showing significantly longer survival. Analysis of the relationship between the number of Ra-223 injections and OS revealed that (c) patients receiving >3 injections had significantly better survival outcomes, and (d) this benefit was even more significant in patients with previous chemotherapy. CI: confidence interval; Ra-223: radium 223; HR: hazard ratio.

Supplementary Table 2.

Summary of the overall survival rates in all patients and subgroups

	Median OS and 95% CI (months)	6-month OS rate and 95% CI	12-month OS rate and 95% CI	18-month OS rate and 95% CI	24-month OS rate and 95% CI
Total patients (n=93) (%)	14.1 (12.0–16.2)	81.7 (73.5–90.9)	59.3 (48.9–71.8)	40.2 (29.1–55.4)	23.0 (12.2–43.6)
Patients without previous chemotherapy (n=42) (%)	18.7 (12.3–25.1)	82.5 (70.7–96.3)	75.4 (61.7–92.0)	56.3 (39.7–80.0)	29.2 (11.4–74.8)
Patients with previous chemotherapy (n=51) (%)	8.9 (3.1–14.7)	81.4 (70.0–93.8)	46.3 (32.8–64.5)	27.5 (15.6–48.7)	18.4 (8.2–41.0)
Patients with ≤3 injections (n=52) (%)	7.8 (5.8–9.8)	66.9 (53.1–84.1)	37.8 (24.6–58.2)	21.6 (10.6–43.9)	21.6 (10.6–43.9)
Patients with >3 injections (n=41) (%)	18.7 (12.3–25.1)	92.5 (84.7–100)	78.8 (66.7–93.2)	58.3 (42.5–79.8)	22.7 (8.2–62.8)
Patients with previous chemotherapy and ≤3 injections (n=34) (%)	6.9 (5.4–8.4)	68.2 (52.2–89.1)	27.5 (14.4–52.5)	13.8 (4.9–38.5)	13.8 (4.9–38.5)
Patients with previous chemotherapy and >3 injections (n=17) (%)	20.6 (11.4–29.8)	94.1 (83.6–100)	75.6 (57.4–99.6)	54.0 (31.4–93.0)	27.0 (8.8–82.8)

OS: overall survival; CI: confidence interval

OS data in different subgroups are shown in Table 2 and Figure 1b–1d. Median OS in subgroups without previous chemotherapy (compared to cases with previous chemotherapy, 18.7 months vs 8.9 months, P = 0.026) and administered >3 Ra-223 injections (compared to ≤3 injections, 18.7 months vs 7.8 months, P = 0.002) were significantly elevated. OS rates in various subgroups at 6 months, 12 months, 18 months, and 24 months are presented in Supplementary Table 2.

Table 2.

Comparative analysis of overall survival in different subgroups

Subgroup	Patient (n)	Overall survival (month), median	95% CI	P-value of the log-rank test between subgroups
Combination therapy				0.162
Ra-223 monotherapy	60	13.6	10.8–16.4
Combination therapy	33	18.7	NA
Ra-223 treatment line				0.525
Treatment line ≤3	45	15.0	10.9–19.1
Treatment line >3	48	13.2	6.6–19.8
Previous chemotherapy				0.026
No	42	18.7	12.3–25.1
Yes	51	8.9	3.1–14.7
Number of Ra-223 injections				0.002
≤3 injections	52	7.8	5.8–9.8
>3 injections	41	18.7	12.3–25.1

Ra-223: radium 223; CI: confidence interval; NA: not available

Subgroup analysis using the Kaplan–Meier survival curves of patients administered androgen receptor pathway inhibitors (ARPIs) and abiraterone is shown in Supplementary Figure 1 ^{(63.5KB, tif)} . Median OS was not reached in the ARPI group, compared to 12.8 (95% CI: 9.2–not available) months in the abiraterone group.

Univariable analyses were conducted on the above factors. As shown in Table 3, previous chemotherapy and >3 Ra-223 injections were significantly associated with OS (P = 0.029 and 0.003, respectively). Multivariable analysis showed that >3 Ra-223 injections was an independent protective factor of OS (hazard ratio [HR] = 0.443, 95% CI: 0.234–0.837; P = 0.012) after adjusting for age, combination therapy, and previous chemotherapy (Table 3).

Table 3.

Univariable and multivariable analyses of overall survival

Influencing factor	Univariable		Multivariable
	HR (95% CI)	P	HR (95% CI)	P
Age ≥70 years	1.040 (0.571–1.880)	0.900	1.395 (0.745–2.610)	0.299
Combination therapy	0.637 (0.337–1.200)	0.170	0.818 (0.412–1.620)	0.565
Using bone protectants	0.769 (0.427–1.390)	0.380	-	-
Treatment line ≤3	0.824 (0.456–1.490)	0.520	-	-
Previous chemotherapy	2.010 (1.070–3.760)	0.029	1.642 (0.845–3.190)	0.143
Total number of Ra-223 injections >3	0.398 (0.219–0.726)	0.003	0.443 (0.234–0.837)	0.012

Ra-223: radium 223; HR: hazard ratio; CI: confidence interval; -: not applicable?

Effect of Ra-223 on pain relief

Of the 93 patients, 33.3% used pain-control medications during the first injection. This proportion decreased with the increasing number of injections (Supplementary Figure 2 ^{(145.8KB, tif)} ). After the third injection, 13.2% of the examined patients were administered pain-control medications, and the proportions of patients reporting a pain score of 3 or 4 decreased (Table 4). After six injections, only 4.2% of patients reported a pain score of 3, with the majority of cases (62.5%) having almost no pain (score 0) and only 8.2% using pain control medications.

Table 4.

Changes in pain grades

Pain grade	Baseline (n=93)	After injection 1 (n=93)	After injection 2 (n=62)	After injection 3 (n=50)	After injection 4 (n=40)	After injection 5 (n=29)	After injection 6 (n=24)
0	13 (14.0)	21 (22.6)	22 (35.5)	22 (44.0)	22 (55.0)	20 (69.0)	15 (62.5)
1	24 (25.8)	29 (31.2)	22 (35.5)	18 (36.0)	10 (25.0)	4 (13.8)	5 (20.8)
2	26 (28.0)	28 (30.1)	11 (17.7)	8 (16.0)	8 (20.0)	5 (17.2)	3 (12.5)
3	29 (31.2)	14 (15.1)	6 (9.7)	1 (2.0)	0 (0)	0 (0)	1 (4.2)
4	1 (1.1)	1 (1.1)	1 (1.6)	1 (2.0)	0 (0)	0 (0)	0 (0)

All data are shown as n (%)

ALP levels

As shown in Supplementary Table 3, median ALP showed a significant downward trend during the six-injection treatment period. The proportion of patients with >3 injections experiencing a decrease in ALP was higher than that of patients with ≤3 injections (65.9% vs 46.2%, P = 0.041).

Supplementary Table 3.

Changes in alkaline phosphatase

Parameter	Baseline	After injection 1	After injection 2	After injection 3	After injection 4	After injection 5	After injection 6
ALP (IU/l), median (Q1–Q3)	146 (102–228)	156 (91.0–236)	112 (83.0–187)	119 (79.0–163)	103 (76.0–155)	89.0 (75.0–118)	89.0 (74.8–125)

ALP: alkaline phosphatase

Safety

As shown in Table 5, the most common adverse event was anemia (30.1%), mostly grade I (12.9%), followed by thrombocytopenia (15.1%), nausea (9.7%), weakness (8.6%), and decreased white blood cells (7.5%). Grade III and IV anemia were detected in 6.4% and 2.2% of patients, respectively, and grade III thrombocytopenia in 2.2% of patients. Only three patients with grade III or IV anemia required transfusion therapy. During treatment, one pathologic fracture and one spinal cord compression were observed.

Table 5.

Summary of adverse events (total patients=93)

Adverse event	Patients, n (%)
Nausea	9 (9.7)
Vomiting	6 (6.5)
Diarrhea	2 (2.2)
Weakness	8 (8.6)
Loss of appetite	4 (4.3)
Constipation	2 (2.2)
Gastrointestinal discomfort	1 (1.1)
Anemia	28 (30.1)
Grade I	12 (12.9)
Grade II	7 (7.5)
Grade III	6 (6.5)
Grade IV	2 (2.2)
Grade data missing	1 (1.1)
Decreased white blood cells	7 (7.5)
Grade I	3 (3.2)
Grade II	1 (1.1)
Grade data missing	3 (3.2)
Thrombocytopenia	14 (15.1)
Grade I	4 (4.3)
Grade II	1 (1.1)
Grade III	2 (2.2)
Grade data missing	7 (7.5)

DISCUSSION

This single-center, retrospective, real-world study comprehensively examined the effectiveness and safety of Ra-223 in patients with mCRPC and bone metastasis in real-world clinical practice in China. The above results suggest that Ra-223 shows favorable effectiveness in mCRPC patients with bone metastases, with a median OS of 14.1 months. The number of injections was independently associated with OS, with those administered >3 Ra-223 injections achieving a better OS than those with ≤3 injections. In addition, Ra-223 effectively relieved pain and decreased ALP levels, with a favorable safety profile. Nevertheless, less than half of patients received >3 injections, and very few completed the six-injection regimen, emphasizing the importance of completing the Ra-223 regimen.

The pivotal phase III ALSYMPCA randomized controlled trial showed that compared with placebo, Ra-223 improved OS in patients with mCRPC (14.9 months vs 11.3 months).7 A recent interim analysis of the prospective Ra-223 dichloride in routine clinical practice settings in Taiwan (China, RAPIT) study reported a real-world median OS of 15.7 months in patients with mCRPC administered Ra-223.24 The real-world Ra-223 Alpha Emitter Agent in Non-intervention Safety Study in mCRPC popUlation for Long-teRm Evaluation (REASSURE) study examined patients with mCRPC treated with Ra-223 followed by a taxane (docetaxel or cabazitaxel), reporting a median OS of 24.3 months from the start of Ra-223 treatment,25 highlighting the promising potential of Ra-223 followed by chemotherapy. In the present study, 40.9% of the evaluated patients had an ECOG PS ≥2, and 51.6% received Ra-223 as the third or later-line of therapy. In addition, our patient population had baseline prostate-specific antigen (PSA) levels overtly higher than reported in previous studies (188 ng ml⁻¹ vs 146 ng ml⁻¹ in ALSYMPCA and 59 ng ml⁻¹ in REASSURE). Therefore, there are significant differences in baseline patient characteristics and treatment patterns in the real-world compared to clinical trials. Moreover, variations across various practice settings are also considerable, which would lead to differences in treatment outcomes. Although the baseline condition of the patients included in the present real-world study may be less favorable, some cases received some novel combination therapies (e.g., abiraterone, sipuleucel-T, and enzalutamide), which may also explain why OS could reach 14.1 months.

The findings of Pain Evaluation in Radium-223 (Xofigo^®) Treated mCRPC Patients With Bone Metastases (PARABO), a prospective observational study primarily focused on pain outcomes, indicated that pain was reduced in patients treated with Ra-223, particularly in those administered 5 or 6 injections.26 This is consistent with the current study, in which no cases who completed the six Ra-223 injections had a pain score of 4, only 4.2% of patients had a pain score of 3, and in total, 62.5% of patients had almost no pain (0 point). However, although the conclusion of our study is consistent with PARABO, several methodological factors in study design and patient selection hindered the ability to compare pain results in both studies. Firstly, differences in the pain assessment tools used made quantitative comparisons difficult. As our study was retrospective and used existing clinical data, pain assessment was based on physician-recorded 5-point Visual Analog Scale scores, which are simple and fast to obtain and are suitable for rapid clinical recording. PARABO, on the other hand, was prospective and therefore allowed for the use of the more comprehensive Brief Pain Inventory-Short Form (BPI-SF) for multidimensional pain assessment. Another difference is that our study also used the frequency of analgesic medication as an indicator of pain severity assessment. Analgesic use can relieve pain, making it an important confounding factor in pain assessment. Using it as an outcome enables a more comprehensive evaluation of pain relief from a different perspective. Furthermore, in PARABO, the number of Ra-223 injections was significantly higher than the maximum of 6 injections used in this study. Beyond regional practice patterns, economic conditions, and population characteristics, patient selection bias should also be considered, as prospective studies tend to enroll more compliant patients, while our retrospective design better reflects real-world clinical practice.26

Another intriguing phenomenon affecting the outcome in this study is that less than half of patients were administered >3 injections, and very few completed the six-injection regimen, which is significantly lower than those reported in previous studies (26.9% vs 63% in ALSYMPCA and 80.5% in the J-RAP-BSI trial27). Further, the number of injections was found to be an independent prognostic protective factor in this study. Early treatment discontinuation in this patient subgroup was attributed to various clinical and non-clinical factors. In the present study, the predominant reasons attributed to Ra-223 discontinuation included disease progression, death during treatment, and hematologic adverse reactions. The adverse reactions were likely related to the poorer baseline conditions of patients in the real world, who might have a lower tolerance to treatment. Furthermore, other factors leading to early discontinuation and non-adherence were identified, including perceived lack of effectiveness, financial constraints, and pain flare, which reflects real-world clinical practice. In the J-RAP-BSI trial, reasons for discontinuation before completing the 6 doses included progression (11.2%), hematologic toxicity (2.4%), declining ECOG (2.9%), bone pain (2.4%), and patient’s wishes (0.5%),27 suggesting that besides progression, patient tolerance also affects the number of treatment doses administered. Patients with a higher burden of previous treatments in this study may also contribute to lower tolerance to subsequent Ra-223 treatments. Consistently, in a study by Saad et al.,28 the proportion of patients with ≤1 previous cancer treatment was higher among cases administered 5 or 6 Ra-223 doses compared with those who received 1–4 injections. In addition, the coronavirus disease 2019 (Covid-19) epidemic in the surveyed area of this study severely impacted accessibility to regular Ra-223 treatment. Based on our estimates, at least 25 (26.9%) patients were administered irregular or incomplete treatment regimens during both the isolation and disease-outburst periods. These findings underscore the importance of regular treatment protocols in mCRPC management.

The combination therapy of Ra-223 and other drugs is an important treatment strategy in clinical practice. In a study by Sartor et al.,29 Ra-223 showed acceptable safety regardless of concurrent abiraterone or enzalutamide. The Evaluation Combining Radium-223 Dichloride and Abiraterone Acetate (eRADicAte) trial reported clinically meaningful improvement in the quality of life and pain of patients administered Ra-223 with abiraterone acetate plus prednisone.30 The present study also found a certain difference in survival curves between the combination therapy and monotherapy subgroups, indicating a trend toward improved OS with combination therapy. Besides, previous chemotherapy was significantly associated with worse OS in univariable analysis. In the present study, many patients with previous chemotherapy were administered Ra-223 as a later-line treatment, and fewer cases received combination therapy. For patients with previously failed chemotherapy, whether trying docetaxel again or using other therapies is a challenging decision. Patients may have difficulty tolerating chemotherapy or other therapies, leading to insufficient effectiveness. McKay et al.31 indicated that using Ra-223 before chemotherapy could be more helpful in completing the Ra-223 treatment course. Thus, the use of Ra-223 as a frontline treatment of mCRPC should be given critical consideration, which may confer more benefits.

The safety of Ra-223 has been evaluated in other studies, suggesting favorable patient tolerability and manageable toxicity.7,8,9,32 Ra-223 concentrates in areas with high bone turnover, and the range of α radiations is only about 100 µm.6 Attention is also paid to other safety concerns, e.g., development of sarcopenia, in mCRPC patients treated with Ra-223.33 These authors showed that Ra-223 does not accelerate sarcopenia, with the sarcopenia observed in patients treated with mCRPC resulting from factors other than Ra-223.34 Furthermore, as a radioisotope, Ra-223 has raised potential concerns about radiation, particularly for others around the patients. Actually, the external radiation produced from excreted Ra-223 would be below 1 mSv, which is lower than the background radiation exposure, suggesting no exposure issue for people living with the patient and the medical personnel.35,36 The adverse reactions of Ra-223 in the present real-world study were generally similar to those reported in previous studies, with no indication of any additional safety concerns. Hematologic adverse reactions are one of the main factors affecting drug discontinuation, and monitoring and timely intervention should be strengthened in clinical practice.

The present study also had several limitations. The retrospective design is associated with potential selection bias, and the data analyzed were limited to patient charts. Important data, such as subsequent treatment options after progression on Ra-223 were lacking, which should be addressed in further prospective studies. Additionally, this was a single-center study, and the generalizability of the obtained findings should be further verified. Finally, follow-up was relatively short, and further follow-up is needed for longer-term OS evaluation in the future.

This retrospective real-world study suggests that Ra-223 has favorable effectiveness in Chinese patients with mCRPC and bone metastases. The survival benefits of Ra-223 may increase with the number of Ra-223 injections and completion of the six-injection treatment. Attention should be paid to monitoring hematologic toxicity and completing the number of injections to achieve an optimal therapeutic effect. Combination therapy and frontline treatment could also be potential ways for Ra-223 therapy to improve patient prognosis.

AUTHOR CONTRIBUTIONS

DWY, CL, HKW, and ZH had full access to all of the data in the study and take responsibility for the integrity of the data and the accuracy of the data analysis. BD and YZ were involved in the study concepts and design. HXJ, CL, YBD, HXJ, YW, HKW, XJQ, and GWL were involved in the acquisition, analysis, and interpretation of data. HKW, CL, and ZH supervised the analysis. HKW drafted the manuscript. All authors read, critically revised, and approved the final manuscript.

COMPETING INTERESTS

YBD, HXJ, and YW are the employees of the Medical Affairs Department, Bayer Healthcare. While Bayer Healthcare participated in scientific discussions regarding the study, the company had no role in the study design, data collection, statistical analysis, interpretation of the data, or the decision to submit the manuscript. The authors Bo Dai and Ding-Wei Ye are members of the Editorial Board of Asian Journal of Andrology, but they have not been involved in the peer-review or decision-making process for this manuscript. The remaining authors declare no competing interests.

Supplementary Figure 1

Kaplan–Meier curves depicting OS in patients receiving Radium-223 in combined with other ARPIs (excluding Abiraterone) versus Abiraterone. ARPIs, androgen receptor pathway inhibitors.

Supplementary Figure 2

(a) Distribution of the pain score changes. (b) Waterfall plot of maximum pain score reduction compared with baseline.