CADTH Health Technology Review

Treatment Sequences of Androgen Receptor–Targeted Agents for Prostate Cancer

Rapid Review

Authors: Khai Tran, Sarah McGill

Key Messages

• No evidence was found on the treatment sequences of androgen receptor–targeted agents in patients with castration-sensitive prostate cancer.

• Evidence from retrospective studies, including those within a systematic review, suggests that sequential treatment of abiraterone followed by enzalutamide is more favourable than enzalutamide followed by abiraterone in improving clinical outcomes such as response rate and progression-free survival, but not overall survival, in patients with castration-resistant prostate cancer.

• Evidence from a retrospective study suggests that docetaxel-containing treatment sequences with androgen receptor–targeted agents may improve progression-free survival compared to sequential therapy with androgen receptor–targeted agents alone in patients with castration-resistant prostate cancer.

• Evidence from a retrospective study did not reveal differences in clinical outcomes of patients with castration-resistant prostate cancer treated with sequential androgen receptor–targeted agents with or without interposed chemotherapy or radium-223.

• These findings were in line with those observed in a 2019 CADTH report.¹ However, the findings should be interpreted with caution due to low-quality evidence.

• No comparative cost-effectiveness studies were identified.

Context and Policy Issues

Prostate cancer is the most common cancer among men in Canada.² Approximately 1 in 9 Canadian men will be diagnosed with prostate cancer in their lifetime.² If prostate cancer is detected early, the survival rate is close to 100%; however, 3 of 4 men will die if the cancer is detected late.² The death rate from prostate cancer has dropped by 50% since 1994.²

In the early development of prostate cancer, cancer cells require androgens to grow; therefore, the cancer at this stage is called castration-sensitive prostate cancer (CSPC).³ Androgen deprivation therapies (ADTs) are treatments that reduce androgen production by the testicles through a surgical procedure or gonadotropin-releasing hormone agonists or antagonists.³ However, patients treated with an ADT that blocks androgen production eventually become resistant, meaning that the cancer can progress even when the levels of androgens in the body are very low or undetectable.³ The cancer at this stage is termed castration-resistant prostate cancer (CRPC). Alternative hormone therapies for prostate cancer include treatments that block the action of androgens or their synthesis.³ Blocking the action of androgens can be achieved by using an androgen receptor blocker, such as enzalutamide (ENZ), apalutamide, or darolutamide, which competes with androgens for binding to androgen receptors, inhibiting the action of androgens in stimulating cancer cell growth.³ Abiraterone acetate (AA) is an androgen biosynthesis inhibitor that blocks the enzyme CYP17.³ Both AA and ENZ are androgen receptor–targeted agents (ARTAs). Various treatment options for CRPC include complete blocking of androgen action using ENZ + ADT, complete inhibition of androgen production using AA and ADT, immunotherapy using a cell-based immunotherapy named Sipuleucel-T, chemotherapy using taxanes such as docetaxel (DTX) or cabazitaxel (CTX), radiopharmaceutical radium-223, and sequential therapies among ARTAs and taxanes.⁴

A recent CADTH report¹ identified weak evidence on 4 sets of comparative sequences of ARTA in the treatment of patients with metastatic CRPC. These were ENZ-to-ENZ + AA versus ENZ-to-AA, AA-to-ENZ versus ENZ-to-AA, ARTA-to-ARTA versus ARTA-to-taxane (DTX or CTX), and ARTA-to-ARTA versus ARTA-to-taxane-to-ARTA. It was found that AA had modest benefit in patients with metastatic CRPC (mCRPC) who progressed after treatment with ENZ, and the combination of ENZ + AA given as second-line therapy was not indicated because of observed adverse effects. The AA-to-ENZ sequence provided better clinical outcomes than the ENZ-to-AA sequence in mCRPC patients. After ARTA therapy, patients with mCRPC who received second-line taxane therapy achieved higher response rates compared with second-line ARTA. There was no difference in efficacy between ARTA-to-ARTA and ARTA-to-taxane-to-ARTA. No comparative cost-effectiveness studies were identified in the previous CADTH report.¹

The aim of this report is to update and extend the previous report by reviewing the comparative clinical effectiveness and cost-effectiveness of varying treatment sequences of ARTA in patients with prostate cancer, including CSPC and CRPC with or without metastasis.

Research Questions

1. What is the comparative clinical effectiveness of various treatment sequences of ARTAs in patients with prostate cancer?

2. What is the comparative cost-effectiveness of various treatment sequences of ARTAs in patients with prostate cancer?

Methods

Literature Search Methods

A limited literature search was conducted by an information specialist on key resources including MEDLINE, the Cochrane Database of Systematic Reviews, the international health technology assessment database, the websites of Canadian and major international health technology agencies, as well as a focused internet search. The search strategy comprised both controlled vocabulary, such as the National Library of Medicine’s MeSH (Medical Subject Headings), and keywords. Part of the search is an update of a literature search strategy developed for a previous CADTH report¹ on ARTA drugs for CRPC. The initial search was limited to English-language documents published between January 1, 2014, and May 10, 2019. For the current report, database searches were rerun on February 1, 2021, to capture any articles published since the initial search date. The search of major health technology agencies was also updated to include documents published since May 2019.

An additional search was conducted with the main search concepts of ARTAs (including berdazimer sodium, ralaniten acetate, enzalutamide, apalutamide, abiraterone, darolutamide, Proxalutamide, and seviteronel) and CSPC. The additional search was also limited to English-language documents published between January 1, 2017 and February 1, 2021. Search filters were applied to limit retrieval to health technology assessments, systematic reviews (SRs), meta-analyses, or network meta-analyses, randomized controlled trials, controlled clinical trials, or any other type of clinical trial, and economic studies. Comments, newspaper articles, editorials, and letters were excluded. Where possible, retrieval was limited to the human population.

Selection Criteria and Methods

One reviewer screened citations and selected studies. In the first level of screening, titles and abstracts were reviewed and potentially relevant articles were retrieved and assessed for inclusion. The final selection of full-text articles was based on the inclusion criteria presented in Table 1.

Exclusion Criteria

Articles were excluded if they did not meet the selection criteria outlined in Table 1, were duplicate publications, or were published before 2017. SRs in which all relevant studies were captured in other more recent or more comprehensive SRs were excluded. Primary studies retrieved by the search were excluded if they were captured in 1 or more included SRs or if they were included in the recent CADTH report.¹

Critical Appraisal of Individual Studies

The included publications were critically appraised by 1 reviewer using the following tools as a guide: A MeaSurement Tool to Assess systematic Reviews 2 (AMSTAR 2)⁵ for SRs, and the Downs and Black checklist⁶ for non-randomized studies. Summary scores were not calculated for the included studies; rather, the strengths and limitations of each included publication were described narratively.

Summary of Evidence

Quantity of Research Available

A total of 638 citations were identified in the literature search. Following screening of titles and abstracts, 621 citations were excluded and 17 potentially relevant reports from the electronic search were retrieved for full-text review. No potentially relevant publications were retrieved from the grey literature search for full-text review. Of these potentially relevant articles, 14 publications were excluded for various reasons, and 3 publications met the inclusion criteria and were included in this report. These comprised 1 SR and 2 non-randomized studies. Appendix 1 presents the PRISMA⁷ flow chart of the study selection.

Summary of Study Characteristics

The detailed characteristics of the included SR⁸ (Table 2) and primary studies^9,10 (Table 3) are presented in Appendix 2.

Study Design

The included SR⁸ included 10 studies (9 retrospective and 1 prospective), 8 of which were included in the meta-analysis. Of the included studies in this SR, 5 retrospective studies had been identified in the previous CADTH report.¹ The literature search was performed from 4 main databases from database inception to December 2019. The quality of the included studies was assessed using the Newcastle-Ottawa Scale. This instrument contains 8 items within 3 domains, with a total maximum score of 9 (4 points for Selection, 2 points for Comparability, and 3 points for Outcomes).¹¹ Random-effects and fixed-effects model meta-analyses were used to provide the summary estimates. Publication bias was assessed using funnel plots.

Both of the included primary studies^9,10 were retrospective in design: 1 was a medical record review⁹ and the other was a registry review.¹⁰ One study⁹ used propensity score matching to minimize selection bias from potential confounders. The other study¹⁰ addressed confounders by excluding patients with missing data from the multivariate analysis.

Country of Origin

The included SR⁸ was conducted by authors from various countries including Canada. The primary studies were conducted by authors from Japan⁹ and the Netherlands.¹⁰

Patient Population

The SR⁸ included studies involving CRPC patients treated with sequential therapy of 2 ARTA agents: AA and ENZ. Patients’ median ages ranged from 62 years to 78 years. Most patients had a Gleason score of 8 of less, meaning their cancers were likely to spread more rapidly.¹² Sample sizes of the included studies ranged from 35 to 255 patients, with a total of 1,096 patients.

The retrospective medical record review study⁹ included 315 DTX/ARTA-naïve CRPC patients who were treated with first-line ARTA (i.e., AA or ENZ) or DTX followed by second-line DTX or ARTA. Patients’ ages ranged from 67 years to 82 years, and approximately half of the patients had a Gleason score of 9 or greater.

The retrospective registry review study¹⁰ included 273 CRPC patients who were treated with a first-line ARTA such as AA + prednisone (P) or ENZ followed by a second-line ARTA such as ENZ or AA + P (ARTA-to-ARTA) or a first-line ARTA followed by a life-prolonging drug (LPD) such as DTX, CTX, or Ra-223 followed by an ARTA (ARTA-to-LPD-to-ARTA). Patients’ ages ranged from 50 years to 90 years, and approximately half of the patients had a Gleason score of between 8 and 10.

Interventions and Comparators

The SR⁸ compared the effect of ARTA sequential therapy and had 2 groups of patients: those treated with AA followed by ENZ (AA-to-ENZ) and those treated with ENZ followed by AA (ENZ-to-AA).

The retrospective medical record review study⁹ compared the effect of ARTA (AA or ENZ) sequential therapy with DTX-containing sequential therapy. There were 2 main groups of patients: those treated with ARTA followed by ARTA (ARTA-to-ARTA), and those treated with ARTA followed by DTX (ARTA-to-DTX) or DTX followed by ARTA (DTX-to-ARTA). The study also conducted subgroup analysis of AA-to-ENZ versus ENZ-to-AA.

The retrospective registry review study¹⁰ had 2 main groups of patients: ARTA-to-ARTA and ARTA-to-LPD-to-ARTA. The ARTA could be either AA + P or ENZ. If patients were treated with AA + P as the first ARTA treatment sequence then the second ARTA sequence was ENZ (AA + P-to-ENZ) or vice versa (ENZ-to-AA + P). The study also conducted subgroup analysis of AA + P-to-ENZ versus ENZ-to-AA + P.

Outcomes

Clinical outcomes considered in the SR⁸ were combined progression-free survival (PFS), prostate-specific antigen (PSA) PFS (PSA-PFS), overall survival (OS), and PSA response rate. Definitions of the outcomes were not provided in the SR.

The primary outcome in the retrospective medical record review study⁹ was OS, which was defined as the period from the start of first-line treatment after resistance to castration to death of any cause. The secondary outcome was PFS, which was defined as the period from the start of first-line treatment after resistance to castration to progression on second-line treatment. The diagnosis of progression in this study was largely based on PSA progression, but not based on radiographic or clinical assessment. It was therefore classified as PSA-PFS.

The retrospective registry review study¹⁰ considered PSA response to ARTA2 as the primary outcome, which was defined as a PSA decline of 50% or greater from baseline. The secondary outcome in this study was treatment duration, which was calculated as the interval between start and stop of the second ARTA.

Summary of Critical Appraisal

The detailed quality assessments of the included SR⁸ (Table 4) and primary studies^9,10 (Table 5) are presented in Appendix 3.

The included SR⁸ was explicit in its objective and inclusion criteria for the review, selection of study design for inclusion, and comprehensive literature search strategy. It was unclear if the study selection was performed in duplicate; however, 2 reviewers independently extracted and assessed the quality of the included studies. The SR did not report whether or not a protocol had been published before the conduct of the review. The SR did not report the sources of funding of the studies included in their review nor provide a list of excluded studies. The SR described the characteristics of the included studies in adequate detail, used appropriate techniques to assess the quality of the included studies (i.e., the Newcastle-Ottawa Scale), used appropriate methods to combine the results (i.e., fixed-effects or random-effects model meta-analyses). The authors did not perform subgroup analyses based on the quality of the included studies nor account for the risk of bias in individual studies when interpreting or discussing the results. Publication bias was assessed using funnel plots. The authors declared that they had no conflict of interest for conducting the review. Overall, the included SR was of acceptable methodological quality.

In terms of reporting, both of the included retrospective studies^9,10 clearly described the objectives, the main outcomes measured, the characteristics of the included patients, and the main findings. However, treatment dosage of the interventions was not reported, and adverse events were not considered as outcomes. In terms of external validity, it was unclear if the participants in both studies were representative of the entire populations from which they were selected. In terms of internal validity, there was a risk of selection bias as both studies had a retrospective design and the method of selection was not clearly described. Both studies used appropriate statistical methods in the analyses, and potential confounders were controlled using appropriate approaches. In both studies, the sample sizes were not based on power calculations; therefore, it remains unclear if the studies were adequately powered to detect clinically meaningful differences. Overall, the methodological quality of the included primary studies was considered low.

Summary of Findings

Clinical Effectiveness of Treatment Sequences of ARTA

The main findings and authors’ conclusions of the SR⁸ and the primary clinical studies^9,10 are presented in Appendix 4.

Combined PFS

The meta-analysis of data from 4 studies in the SR⁸ showed that the sequential treatment of AA-to-ENZ was significantly associated with better combined PFS in patients with CRPC compared with ENZ-to-AA. The overall pooled hazard ratio (HR) was 0.62 (95% confidence interval [CI], 0.49 to 0.78; P < 0.001).

PSA-PFS

The meta-analysis data from 3 studies (1 RCT and 2 retrospective studies) in the SR⁸ showed that the sequential treatment of AA-to-ENZ was significantly associated with better PSA-PFS in patients with CRPC compared with ENZ-to-AA. The overall pooled HR was 0.48 (95% CI, 0.36 to 0.61; P < 0.001).

The retrospective medical record review study⁹ found that sequential treatment of ARTA-to-ARTA was associated with significantly shorter PSA-PFS compared with DTX-containing sequential therapy (i.e., DTX-to-ARTA or ARTA-to-DTX). Median PFS values were 13.7 and 18.9 months, respectively (HR = 1.65; 95% CI, 1.23 to 2.17; P < 0.001). Analysis of the propensity score–matched cohort also revealed that PSA-PFS was significantly shorter in the ARTA-to-ARTA group compared with the DTX-containing group (median PFS: 12.9 months versus 21.6 months; HR = 1.70; 95% CI, 1.16 to 2.48; P = 0.07). Subgroup analysis showed that the significant difference was mainly attributed by the difference between ARTA-to-ARTA and DTX-to-ARTA (median PFS: 13.7 months versus 22.7 months; HR = 1.92; 95% CI, 1.42 to 2.60; P < 0.001), but not between ARTA-to-ARTA and ARTA-to-DTX (median PFS: 13.7 months versus 15.6 months; HR = 1.15; 95% CI, 0.81 to 1.62; P = 0.435). Subgroup analysis also showed no significant difference in PFS between AA-to-ENZ and ENZ-to-AA groups (median PFS: 14.3 months versus 12.4 months; HR = 0.92; 95% CI, 0.59 to 1.44; P = 0.724).

OS

The meta-analysis data from 5 studies in the SR⁸ showed that the sequential treatment of AA-to-ENZ was not significantly associated with better OS in patients with CRPC compared with ENZ-to-AA. The overall pooled HR was 0.77 (95% CI, 0.59 to 1.01; P = 0.055).

The retrospective medical record review study⁹ found no significant difference in OS between the ARTA-to-ARTA group and the DTX-containing group (median OS: 34.6 months versus 38.7 months; HR = 1.05; 95% CI, 0.74 to 1.50; P = 0.781). Results in the propensity score–matched cohort also revealed no significant difference in OS between groups (median OS: 37.9 months versus 45.5 months; HR = 1.10; 95% CI, 0.68 to 1.79; P = 0.701). Subgroup analysis also showed no difference in OS between ARTA-to-ARTA and DTX-to-ARTA (median OS: 34.6 months versus 43.5 months; HR = 1.24; 95% CI, 0.84 to 1.81; P = 0.28) or between ARTA-to-ARTA and ARTA-to-DTX (median OS: 34.6 months versus 24.2 months; HR = 0.71; 95% CI, 0.45 to 1.13; P = 0.150). There was also no significant difference in OS between AA-to-ENZ and ENZ-to-AA groups (median OS: 31.9 months versus 42.0 months; HR = 0.62; 95% CI, 0.33 to 1.14; P = 0.123).

PSA Response Rate

The meta-analysis of data from 6 studies in the SR⁸ showed that the PSA response rate of ENZ as the first ARTA agent in ENZ-to-AA sequential therapy was significantly higher than AA as the first ARTA agent in AA-to-ENZ (response rate: 57.3% versus 46.7%; rate ratio [RR] = 1.31; 95% CI, 1.08 to 1.58; P = 0.025). The meta-analysis of data from 7 studies showed that the PSA response rate of AA as the second ARTA agent in the ENZ-to-AA group was also significantly lower than ENZ as second ARTA agent in AA-to-ENZ group (response rate: 8.2% versus 28.9%; RR = 0.30; 95% CI, 0.19 to 0.46; P < 0.001). When both agents were considered together, the pooled PSA response rates from 4 studies were significantly lower with the ENZ-to-AA compared with AA-to-ENZ sequence (response rate: 4.2% versus 23.5%; RR = 0.21; 95% CI, 0.09 to 0.47; P < 0.001).

The retrospective registry review study¹⁰ found no significant difference in PSA response rates of second ARTA between the ARTA-to-ARTA and ARTA-to-LPD-to-ARTA sequences in the overall cohort (20% versus 18%; P = 0.297). When 30% of patients were excluded from multivariate binary logistic regression analysis due to missing PSA response of second ARTA, the results showed no significant difference in the PSA response rate of the second ARTA between the ARTA-to-ARTA and ARTA-to-LPD-to-ARTA sequences (OR = 0.890; P = 0.89). Subgroup analysis also showed no significant difference in the PSA response rates of ARTA between the AA + P-to-ENZ and ENZ-to-AA + P sequences (23% versus 14%; P = 0.159).

Treatment Duration

The retrospective registry review study¹⁰ found that the median treatment durations of the second ARTA in the ARTA-to-ARTA and ARTA-to-LPD-to-ARTA sequences were 3.2 months (interquartile range [IQR], 1.9 to 7.5) and 3.2 months (IQR, 1.8 to 5.9; P = 0.04), respectively. Subgroup analysis showed no significant difference in the second ARTA treatment duration between AA + P-to-ENZ (3.2 months; IQR, 1.9 to 5.9) and ENZ-to-AA + P sequence (3.2 months; IQR, 1.8 to 7.3).

Cost-Effectiveness of Treatment Sequences of ARTA

No comparative cost-effectiveness studies of different treatment sequences of ARTA in patients with prostate cancer were identified; therefore, no summary can be provided.

Limitations

Although the included SR⁸ was well conducted, its included primary studies had several limitations. Most of the studies included in the analysis had a retrospective design, which has a high risk of selection bias. Heterogeneity in patient characteristics existed among the included studies. The proportion of patients with visceral metastases (range = 2.4% to 77.8%) and the follow-up times (range = 13.5 months to 56 months) varied widely among studies. Although the statistical heterogeneity was not significant in the pooled analyses as judged by the chi-square and I² statistics, the clinical heterogeneity varied among studies which could affect the findings. Seven of 10 included studies were from Japan, thus limiting the generalizability of the findings to other populations. However, the findings may be applicable to the Canadian context because 2 of the included studies in the SR were from Canada and the US, and those findings were similar to those of the pooled analyses.

The results of the 2 retrospective non-randomized studies^9,10 could have been affected by selection bias due to the nature of the study design. In addition, certain patient characteristics were imbalanced between groups. In the retrospective medical record review study,⁹ there were significant differences in baseline characteristics between groups, such as median age, proportion of patients who had a history of definitive therapy, serum PSA level, presence of pain, and visceral metastases. In the retrospective registry review study,¹⁰ significant differences in baseline characteristics between groups were median age, visceral metastases, laboratory values, and time of switching between ARTA. Although propensity score matching and multivariate analysis were performed to reduce the effect of selection bias, residual confounders may have not been excluded. Sample sizes were not based on power calculations; therefore, the lack of significant differences in the results between groups may be caused by insufficient sample sizes. Data on treatment decisions were not captured, and the known patient and disease characteristics may play a role in treatment selection by attending physicians, thus having great influence on the findings. Treatment-related adverse events were not reported, therefore the comparative safety among treatment sequences was unclear.

Conclusions and Implications for Decision- or Policy-Making

One SR⁸ and 2 retrospective non-randomized studies^9,10 were included to address the comparative clinical effectiveness of various treatment sequences of ARTA in patients with prostate cancer. All patients in the included studies had CRPC, most of which were metastatic. No studies regarding the comparative cost-effectiveness of treatment sequences of ARTA in patients with prostate cancer were identified.

Three sets of comparative sequences of ARTA in the treatment of patients with mCRPC were identified: AA-to-ENZ versus ENZ-to-AA, ARTA-to-ARTA versus DTX-combined (i.e., ARTA-to-DTX or DTX-to-ARTA), and ARTA-to-ARTA versus ARTA-to-LPD (i.e., DTX, CTX, or Ra-223)-to-ARTA. Results from the SR⁸ suggested that sequential treatment of AA-to-ENZ may offer survival benefits such as combined PFS and PSA-PFS for patients with CRPC compared with the ENZ-to-AA sequence. However, significant differences between groups was not seen in OS. PSA response rates to both first and second ARTA agents were significantly higher in patients treated with an AA-to-ENZ sequence than an ENZ-to-AA sequence. Results from a retrospective study⁹ suggested that a DTX-combined treatment sequence may improve PFS, but not OS, compared with an ARTA-to-ARTA sequence in DTX/ARTA-naïve CRPC patients. Results from another retrospective study¹⁰ showed no difference between sequential ARTA therapies with or without interposed LPD (i.e., DTX, CTX, or Ra-223).

The findings in this updated report were in line with those in the previous CADTH report.¹ First, an AA-to-ENZ sequence appeared to be more favourable than an ENZ-to-AA sequence in patients with mCRPC regarding improved clinical outcomes, such as combined PFS, time to PSA progression, and PSA response rate, but not for OS. Second, an ARTA-to-taxane (i.e., DTX or CTX) sequence appeared to be more effective than an ARTA-to-ARTA treatment sequence in patients with chemotherapy-naïve mCRPC regarding improved clinical outcomes such as combined PFS, time to PSA progression, and PSA response rate. Third, no differences were observed between ARTA-to-ARTA and ARTA-to-taxane-to-ARTA with respect to PSA response rate and time to PSA progression. Considering the limitations, including the retrospective nature of the included studies, collective findings in this report and the previous report¹ should be interpreted with caution.

References

1.Androgen receptor targeted agents for castration resistant prostate cancer: a review of clinical effectiveness and cost-effectiveness (CADTH rapid response report: summary with critical appraisal). Ottawa (ON): CADTH; 2019: https://www.cadth.ca/sites/default/files/pdf/htis/2019/RC1127%20ARTA%20for%20mCRPC%20Final.pdf. Accessed 2021 Feb 2.

2.Canadian Cancer Society. About prostate cancer: statistics. 2020; https://www.prostatecancer.ca/prostate-cancer/about-prostate-cancer/statistics. Accessed 2021 Mar 2.

3.National Cancer Institute. Hormone therapy for prostate cancer. 2019; https://www.cancer.gov/types/prostate/prostate-hormone-therapy-fact-sheet. Accessed 2021 Mar 2.

4.Moussa M, Papatsoris A, Abou Chakra M, Sryropoulou D, Dellis A. Pharmacotherapeutic strategies for castrate-resistant prostate cancer. Expert Opin Pharmacother. 2020;21(12):1431-1448. Medline

5.Shea BJ, Reeves BC, Wells G, et al. AMSTAR 2: a critical appraisal tool for systematic reviews that include randomised or non-randomised studies of healthcare interventions, or both. BMJ. 2017;358:j4008. Medline

6.Downs SH, Black N. The feasibility of creating a checklist for the assessment of the methodological quality both of randomised and non-randomised studies of health care interventions. J Epidemiol Community Health. 1998;52(6):377-384. Medline

7.Liberati A, Altman DG, Tetzlaff J, et al. The PRISMA statement for reporting systematic reviews and meta-analyses of studies that evaluate health care interventions: explanation and elaboration. J Clin Epidemiol. 2009;62(10):e1-e34. Medline

8.Mori K, Miura N, Mostafaei H, et al. Sequential therapy of abiraterone and enzalutamide in castration-resistant prostate cancer: a systematic review and meta-analysis. Prostate Cancer Prostatic Dis. 2020;23(4):539-548. Medline

9.Naito Y, Kato M, Kawanishi H, et al. The clinical benefit of sequential therapy with androgen receptor axis-targeted agents alone in patients with castration-resistant prostate cancer: a propensity score-matched comparison study. Prostate. 2020;80(15):1373-1380. Medline

10.Kuppen MCP, Westgeest HM, van den Eertwegh AJM, et al. Real-world outcomes of sequential androgen-receptor targeting therapies with or without interposed life-prolonging drugs in metastatic castration-resistant prostate cancer: results from the Dutch Castration-resistant Prostate Cancer Registry. Eur Urol Oncol. 2019. Medline

11.Gierish JM, Beadles C, Shapiro A, et al. Appendix B: Newcastle-Ottawa scale coding manual for cohort studies. In: Health disparities in quality indicators of healthcare among adults with mental illness. Washington (DC): Department of Veterans Affairs (US); 2014: https://www.ncbi.nlm.nih.gov/books/NBK299087/#:~:text=The%20Newcastle%2DOttawa%20Scale%20quality,and%203%20points%20for%20Outcomes. Accessed 2021 Mar 2.

12.American Society of Clinical Oncology (ASCO). Prostate cancer: stages and grades. 2020; https://www.cancer.net/cancer-types/prostate-cancer/stages-and-grades. Accessed 2021 Mar 2.

Appendix 1: Selection of Included Studies

Figure 1: Selection of Included Studies

Appendix 3: Critical Appraisal of Included Publications

Table 4: Quality Assessment of the Systematic Review Using the AMSTAR 2 Checklist⁵

AMSTAR 2 = A MeaSurement Tool to Assess systematic Reviews 2; ARTA = androgen receptor–targeted agent; CRPC = castration-resistant prostate cancer; PICO = Population, Intervention, Comparator, Outcomes; RoB = risk of bias.

Table 5: Quality Assessment of Clinical Studies Using the Downs and Black checklist⁶

ARTA = androgen receptor–targeted agent; CI = confidence interval; CRPC = castration-resistant prostate cancer; NA = not applicable.

Appendix 4: Main Study Findings and Authors’ Conclusions

Summary of Findings of the Included Systematic Review

Mori et al. (2020)⁸

Main Study Findings

AA-to-ENZ versus ENZ-to-AA

• Combined PFS (4 studies; N = 395)

  • AA-to-ENZ was significantly associated with better combined PFS in CRPC

  • Pooled HR (95% CI) = 0.62 (0.49 to 0.78); I² = 51.5%; P < 0.001

  • No significant heterogeneity

• PSA-PFS (3 studies; N = 337)

  • AA-to-ENZ was significantly associated with better combined PSA-PFS in CRPC

  • Pooled HR (95% CI) = 0.48 (0.36 to 0.61); I² = 58.3%; P < 0.001

  • No significant heterogeneity

• OS (5 studies; N = 430)

  • ARTA treatment sequence was not significantly associated with OS in CRPC

  • Pooled HR (95% CI) = 0.77 (0.59 to 1.01); I² = 0.0%; P = 0.055

  • No significant heterogeneity

• PSA response rate of first ARTA agent (6 studies; N = 460)

  • 46.7% in AA-to-ENZ and 57.3% in ENZ-to-AA

  • ENZ-to-AA was significantly associated with better PSA response rate of first ARTA agent in CRPC

  • RR (95% CI) = 1.31 (1.08 to 1.58); I² = 0.8%; P = 0.025

  • No significant heterogeneity

• PSA response rate of second ARTA agent (7 studies; N = 606)

  • 28.9% in AA-to-ENZ and 8.2% in ENZ-to-AA

  • AA-to-ENZ was significantly associated with better PSA response rate of second ARTA agent in CRPC

  • RR (95% CI) = 0.30 (0.19 to 0.46); I² = 21.2%; P < 0.001

  • No significant heterogeneity

• PSA response rate of both first and second ARTA agents (4 studies; N = 355)

  • 23.5% in AA-to-ENZ and 4.2% in ENZ-to-AA

  • AA-to-ENZ was significantly associated with better PSA response rate of both first and second ARTA agents in CRPC

  • RR (95% CI) = 0.21 (0.09 to 0.47); I² = 0.0%; P < 0.001

  • No significant heterogeneity

Authors’ Conclusion

“AA-to-ENZ sequential therapy in patients with CRPC was associated with better PFS, PSA-PFS, and PSA response rates. Regardless of sequencing, response to drug therapy was transient for both AA and ENZ when either agent was used as a secondary therapy. Despite this, treatment sequencing is important to achieve the maximum possible benefit from available drugs in CRPC (p. 539).”⁸

Summary of Findings of Included Primary Clinical Studies

Naito et al. (2020)⁹

Main Study Findings

ARTA-to-ARTA (n = 106) versus DTX-combined (i.e., DTX-to-ARTA or ARTA-to-DTX) (n = 209)

• OS in the overall cohort

  • No significant difference between ARTA-to-ARTA and DTX-combined groups

  • Median OS: 34.6 versus 38.7 months

  • HR (95% CI) = 1.05 (0.74 to 1.50); P = 0.781

• OS in the propensity score–matched cohort

  • No significant difference between ARTA-to-ARTA (n = 78) and DTX-combined groups (n = 78)

  • Median OS: 37.9 versus 45.4 months

  • HR (95% CI) = 1.10 (0.68 to 1.79); P = 0.701

• OS in the subset analysis (ARTA-to-ARTA [n = 106] versus DTX-to-ARTA [n = 138])

  • No significant difference between ARTA-to-ARTA and DTX-to-ARTA groups

  • Median OS: 34.6 versus 43.5 months

  • HR (95% CI) = 1.24 (0.84 to 1.81); P = 0.28

• OS in the subset analysis (ARTA-to-ARTA [n = 106] versus ARTA-to-DTX [n = 71])

  • No significant difference between ARTA-to-ARTA and ARTA-to-DTX groups

  • Median OS: 34.6 versus 24.2 months

  • HR (95% CI) = 0.71 (0.45 to 1.13); P = 0.15

• OS in the subset analysis (AA-to-ENZ [n = 54] versus ENZ-to-AA [n = 52])

  • No significant difference between AA-to-ENZ and ENZ-to-AA groups

  • Median OS: 31.9 versus 42.0 months

  • HR (95% CI) = 0.62 (0.33 to 1.14); P = 0.123

• PSA-PFS in the overall cohort

  • ARTA-to-ARTA group was associated with significantly shorter PFS than DTX-combined group

  • Median PFS: 13.7 versus 18.9 months

  • HR (95% CI) = 1.65 (1.23 to 2.17); P < 0.001

• PSA-PFS in the propensity score–matched cohort

  • ARTA-to-ARTA group (n = 78) was associated with significantly shorter PFS than DTX-combined group (n = 78)

  • Median PFS: 12.9 versus 21.6 months

  • HR (95% CI) = 1.70 (1.16 to 2.48); P = 0.007

• PSA-PFS in the subset analysis (ARTA-to-ARTA [n = 106] versus DTX-to-ARTA [n = 138])

  • ARTA-to-ARTA group was associated with significantly shorter PFS than DTX-to-ARTA groups

  • Median PFS: 13.7 versus 22.7 months

  • HR (95% CI) = 1.92 (1.42 to 2.60); P < 0.001

• PSA-PFS in the subset analysis (ARTA-to-ARTA [n = 106] versus ARTA-to-DTX [n = 71])

  • No significant difference between ARTA-to-ARTA and ARTA-to-DTX groups

  • Median PFS: 13.7 versus 15.6 months

  • HR (95% CI) = 1.15 (0.81 to 1.62); P = 0.435

• PSA-PFS in the subset analysis (AA-to-ENZ [n = 54] versus ENZ-to-AA [n = 52])

  • No significant difference between AA-to-ENZ and ENZ-to-AA groups

  • Median PFS: 14.3 versus 12.4 months

  • HR (95% CI) = 0.92 (0.59 to 1.44); P = 0.724

Authors’ Conclusion

“In conclusion, there was no difference in OS outcomes between ARTA-to-ARTA and DTX-combined sequential therapy, while DTX-combined sequential therapy had better PFS outcomes. These data indicate that there are certain patients who will benefit from ARTA-to-ARTA sequential therapy in a daily clinical setting. It is expected that research on additional clinical and molecular predictors of treatment response will be performed (p. 1379).”⁹

Kuppen et al. (2019)¹⁰

Main Study Findings

ARTA1-to-ARTA2 (n = 148) versus ARTA1-to-LPD-to-ARTA2 (n = 125)

• PSA response rate of ARTA2 in the overall cohort

  • No significant differences between ARTA1-to-ARTA2 and ARTA1-to-LPD-to-ARTA2 groups

  • 20% versus 18%; P = 0.297

• PSA response rate of ARTA2 from multivariate binary logistic regression analysis

  • No significant differences between ARTA1-to-ARTA2 and ARTA1-to-LPD-to-ARTA2 groups

  • OR = 0.890; P = 0.89

• PSA response rate of ARTA2 different subgroups

  • No significant differences between AA + P-to-ENZ and ENZ-to-AA + P subgroups

  • 23% versus 14%; P = 0.159

• Treatment duration of ARTA2 in the overall cohort

  • At the end of follow-up, 9% of patients in ARTA1-to-ARTA2 group and 3% of patients in ARTA1-to-LPD-to-ARTA2 group were still on treatment

  • Median (IQR): 3.2 (1.9 to 7.5) versus 3.2 (1.8 to 5.9) months; P = 0.042

• Treatment duration of ARTA2 in different subgroups

  • No significant differences between AA + P-to-ENZ and ENZ-to-AA + P subgroups

  • Median (IQR): 3.2 (1.9 to 5.9) versus 3.2 (1.8 to 7.3) months; P = 0.158

Authors’ Conclusion

“The effect of ARTA2 seems to be low, with a low PSA response rate and a short treatment duration irrespective of interposed chemotherapy or radium-223 (p. 2).”¹⁰

“We observed no differences in outcomes of patients treated with sequential abiraterone acetate plus prednisone and enzalutamide with or without interposed chemotherapy or radium-223 (p. 2).”¹⁰