Visual Abstract
Abstract
Imaging before 223Ra-dichloride (223Ra) therapy is crucial for selecting metastatic castration-resistant prostate cancer (mCRPC) patients with bone-only disease. The purpose of this study was to evaluate if baseline prostate-specific membrane antigen (PSMA) PET/CT (bPSMA) versus CT is associated with outcomes of 223Ra therapy. Methods: A secondary analysis of the data of a prospective observational study (NCT04995614) was performed. Patients received a maximum of 6 223Ra cycles and were retrospectively divided into the bPSMA or baseline CT (bCT) groups. All patients received baseline bone scintigraphy. Primary endpoints were alkaline phosphatase and prostate-specific antigen response. Secondary endpoints were overall survival (OS) and radiologic response. Results: Between 2017 and 2020, 122 mCRPC patients were included: 18 (14.8%) in the bPSMA group and 104 (85.2%) in the bCT group. All baseline characteristics were comparable. No significant differences in alkaline phosphatase or prostate-specific antigen response were found. The bCT group showed an OS significantly shorter than that of the bPSMA group (12.4 vs. 19.9 mo, P = 0.038). In 31 of 76 patients (40.1%) in the bCT group who also received posttherapy CT, lymph node or visceral metastases (soft-tissue involvement [STI]) were detected after 223Ra therapy, compared with 0 of 15 patients in the bPSMA group who received posttherapy PSMA PET/CT or CT. No significant difference in OS was found between patients in the bCT or posttherapy CT subgroup without STI (46/76) and the bPSMA group. Conclusion: bPSMA versus CT does not seem to impact biochemical response during 223Ra therapy in mCRPC patients. Nevertheless, patients in the bCT group had a significantly shorter OS, most likely due to underdetection of STI in this group. Therefore, replacing bCT with PSMA PET/CT appears to be a valuable screening method for identifying patients who will benefit most from 223Ra therapy.
Metastatic castration-resistant prostate cancer (mCRPC) is characterized as prostate cancer progression despite adequate androgen-deprivation therapy (1). Bone metastases are prevalent in over 90% of end-stage mCRPC (2,3) and a leading factor in skeleton-related events (SREs), morbidity, and mortality (4,5). Symptomatic bone metastases can be treated with 223Ra-dichloride (223Ra), a radionuclide that actively incorporates into metastatic lesions, where it emits α-particles that induce tumor cell death (6,7). The phase 3 ALSYMPCA trial demonstrated that treatment with 223Ra compared with placebo significantly prolonged overall survival (OS), led to a more frequent decline in alkaline phosphatase (ALP), enhanced quality of life, and presented fewer SREs (8–11). 223Ra therapy should be withheld in patients with extensive malignant lymphadenopathy or visceral metastases (12,13), of which the latter is observed in up to 32% of mCRPC patients (3). Therefore, baseline imaging plays a crucial role in assessing eligibility for 223Ra therapy and currently consists of either baseline CT (bCT) or prostate-specific membrane antigen (PSMA) PET/CT, in addition to bone scintigraphy (12,14). PSMA PET/CT has a higher diagnostic sensitivity and specificity for detecting pelvic nodal or distant metastases than CT (15–17) and has been shown to be able to detect previously unknown visceral metastases in a small cohort of mCRPC patients who underwent screening for 223Ra therapy (18). To our knowledge, there is only one retrospective study available evaluating the impact of baseline PSMA PET/CT (bPSMA) compared with CT on outcomes of 223Ra therapy, and this study suggested that staging with PSMA PET/CT results in better therapeutic responses with greater declines in ALP and prostate-specific antigen (PSA) because of better patient selection (19). On the basis of the hypothesis that the higher diagnostic accuracy of PSMA PET/CT will lead to better patient selection and therefore outcomes of the proposed treatment, we studied the impact of bPSMA versus CT on outcomes of 223Ra therapy in mCRPC patients.
MATERIALS AND METHODS
Study Design and Patient Population
We performed a secondary analysis of a prospective observational multicenter cohort study that evaluated mCRPC patients treated with 223Ra at 11 institutions throughout The Netherlands between April 2017 and July 2020 (NCT04995614) (20). Eligible patients had histologically proven mCRPC with symptomatic bone metastases and no visceral metastases. The study protocol was approved by the medical ethics committee (CMO 2017-3220) and the institutional review boards of all participating centers. Informed consent was obtained from all individual patients.
Study Procedures and Follow-up
Patients were treated with 223Ra according to the standard of care at a dose of 55 kBq/kg of body weight injected intravenously every 4 wk, with a maximum of 6 injections. Every patient received a baseline bone scintigraphy to determine the extent and localization of bone metastases. Additionally, every patient received a contrast-enhanced high-dose CT of the thorax and abdomen or a low-dose 68Ga-PSMA-11 or 18F-PSMA-1007 PET/CT from the top of the skull down to the mid thigh to detect soft-tissue involvement (STI), defined as lymph or visceral node metastases. The choice for baseline imaging modality was made according to local standard clinical care in the respective hospitals. Baseline imaging was performed no more than 12 wk before the start of therapy. At least 1 wk before every 223Ra injection, laboratory evaluation was performed. Laboratory evaluation and posttherapy imaging (bone scintigraphy combined with either CT or PSMA PET/CT) were performed 4–8 wk after the last injection. All patients were followed until death or March 3, 2023.
Subgroup Categorization
Patients were retrospectively allocated to the bPSMA or bCT group on the basis of imaging modality before 223Ra therapy. Additionally, patients were retrospectively allocated to the posttherapy PSMA PET/CT (pPSMA) or posttherapy CT (pCT) subgroups.
Study Outcomes
The primary endpoint was biochemical response, defined as at least a 30% decline in ALP or PSA level from baseline during 223Ra therapy, as described in the ALSYMPCA trial (9). Any decline in PSA level from baseline during treatment was determined. ALP and PSA superresponders were defined as patients with at least a 50% decline from baseline during treatment.
Secondary endpoints were the best percentage change in ALP or PSA response from baseline during treatment, radiologic response, occurrence of SREs, and OS. Radiologic response after 223Ra therapy was evaluated by monitoring the presence of newly detected STI. Only patients who received identical baseline and posttherapy imaging techniques were eligible for radiologic response evaluation. RECIST version 1.1, the Prostate Cancer Working Group 3 criteria, and the PSMA PET progression criteria were used (21–23). SREs were defined as surgery or radiotherapy to the bone, spinal cord compression, and symptomatic pathologic fractures (22). OS was defined as the time from the start of 223Ra therapy to the date of death from any cause or the date of the last follow-up.
Data Analysis
Statistical analysis was performed for the total cohort and the subgroups. Subgroups were compared using χ2 and Mann–Whitney U tests for categoric and continuous variables, respectively. OS was analyzed for subgroups using Kaplan–Meier curves and the log-rank test. A 2-sided P value of less than 0.05 was considered statistically significant. Statistical analyses were performed using SPSS 27.0 (IBM Corp.), and figures were created using GraphPad Prism 9.0 (GraphPad Software Inc.).
RESULTS
Patient Characteristics
Of the 122 included patients, 18 patients (14.8%) underwent bPSMA and 104 patients (85.2%) underwent bCT imaging (Table 1). All patients received a baseline bone scintigraphy. Seventy-seven patients (63.1%) completed the full course of 6 223Ra cycles. No significant differences in any of the baseline characteristics were found between the 2 groups.
Baseline Patient Demographics and Clinical Characteristics
Biochemical Response
No significant differences in biochemical responses were found between the bPSMA and bCT groups (Table 2). ALP responses were found in 55.6% of bPSMA patients versus 62.5% of bCT patients (P = 0.576). More ALP superresponders were found in the bPSMA group than in the bCT group (50.0% vs. 28.8%, respectively; P = 0.076). In the bPSMA group, more patients had some PSA response than in the bCT group (44.4% vs. 29.8%, respectively; P = 0.219).
Biochemical Response During 223Ra Therapy for Total Cohort and bPSMA and bCT Groups
Radiologic Response
Eleven of 18 bPSMA patients (61.1%) also received a pPSMA (bPSMA/pPSMA subgroup) (Fig. 1). Seventy-six of 104 bCT patients (73.1%) also received a pCT (bCT/pCT subgroup). Patients switched from bPSMA to pCT in 4 of 18 cases (22.2%) and from bCT to pPSMA in 7 of 104 cases (6.7%).
CONSORT (Consolidated Standards of Reporting Trials) diagram. Stratification of study population was based on baseline and posttherapy imaging methods used for radiologic response evaluation.
No significant differences were found in any of the baseline characteristics between the bPSMA/pPSMA and bCT/pCT subgroups (Supplemental Table 1; supplemental materials are available at http://jnm.snmjournals.org). None of the 15 bPSMA patients who received posttherapy imaging were found to have newly detected STI. In 31 of 76 (40.1%) bCT/pCT patients, STI was newly detected after therapy (Table 3), with visceral metastases being detected in 16 patients (51.6%) and lymph node metastases in 20 patients (64.5%). In retrospect, small visceral metastases were already visible on bCT imaging in 2 patients. Among the bCT/pPSMA patients, newly detected STI was found in 6 of 7 patients (85.7%), with visceral metastases being detected in 4 patients (66.7%) and lymph node metastases in 6 patients (100%).
OS Among Subgroups
SREs
SREs occurred during and after 223Ra therapy in 6 of 18 (36.4%) and 30 of 104 (28.8%) patients of the bPSMA and bCT groups, respectively (P = 0.700). External-beam radiotherapy to relieve skeletal symptoms was the most common SRE (61.5%) (Supplemental Table 2).
OS
Median OS for the total cohort was 12.8 mo (95% CI, 3.3–46.4 mo). Patients in the bPSMA group had a significantly longer median OS of 19.9 mo (95% CI, 3.1 mo to not reached) than did those in the bCT group (12.4 mo, 95% CI, 3.3–37.2 mo; P = 0.038) (Fig 2A).
OS among subgroups: OS of bPSMA and bCT groups (A) and OS of bPSMA and bCT/pCT subgroups (B). Latter are divided into further subgroups: without STI, with lymph node metastases, with visceral metastases, and with both lymph node and visceral metastases.
In an exploratory analysis, OS was compared for patients in the bCT/pCT subgroup with and without newly detected STI after therapy. The patients in the bCT/pCT subgroup with newly detected STI had a significantly shorter OS than that in those without newly detected STI (median OS, 10.6 vs. 14.9 mo, respectively; P < 0.01) (Table 3; Fig. 2B). No significant OS difference was found between the bCT/pCT subgroup without newly detected metastases and the overall bPSMA group (P = 0.456).
DISCUSSION
In this prospective cohort study, we investigated the impact of bPSMA versus bCT on the outcomes of 223Ra therapy in mCRPC patients. Patients who received bCT had a significantly shorter OS than those who underwent bPSMA. We did not observe significant differences in biochemical response rates.
Our study found ALP and PSA responses similar to those in the ALSYMPCA trial and previous retrospective studies on 223Ra therapy in mCRPC patients (8,24–26). To date, only Ahmadzadehfar et al. reported outcomes on biochemical response during 223Ra therapy based on different baseline imaging modalities (26). In this study, both ALP and PSA response rates were significantly higher when using bPSMA than when using bCT. Although we did not observe significant differences, the bPSMA group showed more ALP superresponders and PSA responses (any decline) than did the bCT group.
Overall, the OS of 12.8 mo found in our study is in line with other real-world studies evaluating 223Ra therapy (20,27). We found a remarkably longer median OS for patients in the bPSMA group than for patients in the bCT group (19.9 vs. 12.4 mo, respectively), whereas all baseline characteristics were comparable. The most probable explanation for this finding is the absence of newly detected STI in any of the 15 patients in the bPSMA/pPSMA and bPSMA/pCT subgroups, whereas STI was frequently detected after therapy in the bCT/pCT and bCT/pPSMA subgroups (40.1% and 85.7%, respectively). Our exploratory analysis confirmed a significant difference in OS between bCT/pCT patients with and without newly detected STI after therapy. Conversely, the OS of patients in the bCT/pCT subgroup without newly detected STI did not significantly differ from the OS of the bPSMA group. These findings are supported by data in the literature that suggest mCRPC patients with visceral metastases have a significantly shorter OS than patients without visceral metastases (28,29).
It seems unlikely that the newly detected STI after treatment in the bCT group was not already present at the start of 223Ra therapy, since development of visceral disease takes approximately 1.6 y from CRPC diagnosis (3). In retrospect, the posttherapy-detected STI was already visible at the bCT in 2 patients. We encountered a lower proportion of patients with a PSA decline in the bCT group than in the bPSMA group, possibly due to the presence of STI (3). The presumed underdetection of STI at the start of 223Ra therapy in the bCT group might have caused the treatment to be less effective. However, as there was no control group, our results may still be explained by the natural history of the disease and unrelated to 223Ra therapy itself.
Our study has several limitations that should be considered when interpreting the results. The retrospective allocation of patients into the bPSMA and bCT groups might have led to selection bias. However, the groups did not significantly differ in any of the baseline characteristics, and the baseline imaging modality choice was made according to local standards of care in the respective hospitals. To the best of our knowledge, patient characteristics did not influence the imaging modality of choice. In addition, none of the patients received both imaging modalities at baseline. Furthermore, the small sample size of the subgroups may have contributed to the lack of statistical significance in biochemical response analyses. Because of these limitations, conclusions should be drawn with caution.
Our study fills a gap of knowledge as little research has been done on the impact of using PSMA PET/CT instead of CT when selecting mCRPC patients for 223Ra therapy. To our knowledge, our population is the largest cohort of patients addressing this research question. It is probable that the OS benefit associated with PSMA PET/CT as a therapeutic eligibility assessment modality can be extrapolated to other therapies for metastatic prostate cancer. Patients who might benefit less from 223Ra monotherapy because of the presence of STI may be redirected to other treatments, such as the combination of docetaxel plus 223Ra therapy (DORA trial, NCT03574571). Additionally, the outcomes of this study raise the question whether existing evidence-based guidelines are still valid for current daily practice, as PSMA PET/CT appears to become a potential replacement of CT in the management of mCRPC. For instance, a recently published article on the use of 68Ga-PSMA PET/CT for response evaluation of 223Ra therapy demonstrated that total tumor volume within the bone at baseline 68Ga-PSMA PET/CT was associated with treatment response and development of extraosseous disease during treatment (30). This emphasizes the need for prospective randomized clinical trials, preferably incorporating masking of the practitioner, to assess the impact of PSMA PET/CT compared with CT on therapeutic choices and outcomes (31).
CONCLUSION
bPSMA does not appear to be a strong predictor of biochemical response during 223Ra therapy when compared with CT. However, patients who underwent bCT had a significantly shorter OS, most likely due to underdetection of STI before the start of 223Ra therapy. Therefore, replacing bCT with bPSMA appears to be a valuable screening method to identify the patients who will benefit most from 223Ra therapy.
DISCLOSURE
This work was supported by Bayer Healthcare The Netherlands. The funding organization had no role in the design or conduct of the study; collection, management, analysis, or interpretation of the data; or preparation, review, or approval of the article. Dianne Bosch received speaker honoraria from MSD. Wim Oyen received consultancy and speaker fees from Bayer and AAA Novartis. Addy van der Luijtgaarden is a member of the advisory board of the Dutch prostate cancer foundation, received speaker and advisory honoraria from Janssen Pharmaceuticals, and received financial support for an educational program from Astellas. Diederik Somford is a member of the advisory boards of Janssen Pharmaceuticals, Astellas, and Bayer and received research grants from Astellas outside the submitted work. Rick Hermsen was a member of the advisory board of Bayer, received personal fees and travel expenses from Bayer outside the submitted work, and received personal fees from ABX and Blue Earth Diagnostics for image review. Niven Mehra received personal fees from Bayer outside the submitted work, Sanofi, and MSD; received research grants from BMS; and received research grants and personal fees from Pfizer outside the submitted work, AstraZeneca, Astellas, Janssen Pharmaceuticals, MSD, and Roche. Winald Gerritsen is a member of the advisory boards for Bristol-Myers Squibb, Astellas, Janssen Pharmaceuticals, Bayer, Sanofi Genzyme, Amgen, Morphosys, and CureVac; received speaker fees from Bayer and MSD; and received research funding from Astellas, Bayer, and Janssen Pharmaceuticals. Maarten van der Doelen received research grants from Bayer and Janssen Pharmaceuticals and received speaker fees from Astellas. Irma Oving is a member of the advisory boards for MSD, Pfizer, and AstraZeneca and received research grants and personal fees from Astellas, Bayer, Janssen Pharmaceuticals, MSD/AstraZeneca, and Pfizer. No other potential conflict of interest relevant to this article was reported.
KEY POINTS
QUESTION: How does using PSMA PET/CT instead of CT to select patients for 223Ra therapy impact the outcome?
PERTINENT FINDINGS: In this prospective cohort study, we found a significantly longer median OS for patients who received bPSMA compared than for those who received bCT, most likely due to underdetection of STI at the start of 223Ra therapy.
IMPLICATIONS FOR PATIENT CARE: PSMA PET/CT appears to be a valuable screening method to identify the patients who will benefit most from 223Ra therapy.
Footnotes
↵* Contributed equally to this work.
Published online Feb. 29, 2024.
- © 2024 by the Society of Nuclear Medicine and Molecular Imaging.
REFERENCES
- Received for publication September 10, 2023.
- Revision received January 5, 2024.
In this issue
Jump to section
Related Articles
Cited By...
- No citing articles found.