Abstract
Background/Aim: No practical biomarkers predict the response to enzalutamide in chemotherapy-naïve metastatic castration-resistant prostate cancer (mCRPC). The present study aimed to evaluate the prognostic value of the initial-to-nadir prostate-specific antigen (PSA) ratio (I/N PSA) in primary hormone therapy for metastatic hormone-naïve prostate cancer associated with the response to first-line enzalutamide in mCRPC. Patients and Methods: Twenty-eight patients with mCRPC received first-line enzalutamide to determine the associations between I/N PSA in combined androgen blockade and clinical outcomes. The PSA response was defined as ≥90% decline from baseline in patients with mCRPC. Results: The optimal cutoff I/N PSA value for PSA response was 1,219 (sensitivity=71.4%, specificity=92.9%, area under the receiver operating characteristic curve=0.85). The PSA response was 90.9% in the high I/N PSA group and 23.5% in the low I/N PSA group. The median overall survival, prostate cancer-specific survival, and radiographic progression-free survival after initiation of enzalutamide were statistically greater for the high I/N PSA group than the low group. Multivariable analysis showed that I/N PSA was an independent predictor of overall survival (hazard ratio=0.23; p=0.026). Conclusion: In chemotherapy-naïve patients with mCRPC, I/N PSA was a predictive and prognostic biomarker for first-line enzalutamide. The I/N PSA can enable optimization of individual treatment in real-world clinical practice.
- Castration-resistant prostate cancer
- initial-to-nadir prostate-specific antigen ratio
- enzalutamide
- androgen-deprivation therapy
Prostate cancer (PC) is the second most common cancer in males worldwide (1). In 2020, an estimated 1,414,259 males were newly diagnosed with PC, which was approximately 14% of all new cancer cases in men, with 375,304 deaths due to the disease (1). Androgen deprivation therapy (ADT) remains a common treatment option for male patients with metastatic hormone-naïve PC (mHNPC). Despite the fact that PC is highly androgen-dependent and sensitive to hormone therapy, patients with mHNPC treated with ADT (within 18-24 months) inevitably become metastatic castration-resistant PC (mCRPC) (2, 3).
Recently, there has been remarkable advancement in the treatment of metastatic PC. The clinical outcomes for patients with mCRPC, including novel hormonal therapy (NHT) (enzalutamide, abiraterone, apalutamide, and darolutamide), chemotherapy agents (cabazitaxel and docetaxel), immunotherapy, and radium-223, have been improved by multiple agents with diverse mechanisms of action (4). However, the situation has been further complicated because some of these agents have been approved for the treatment of up-front mHNPC combined with ADT. In patients with mHNPC, randomized phase III clinical trials have shown the survival benefits of up-front combination therapy (4). The American and European guidelines recommend the use of up-front combination therapy for all patients with mHNPC, which reflects the current standard of care (5, 6).
However, emerging claims-based data in the up-front era have revealed that patients with mHNPC did not frequently receive the standard of care treatment in real-world settings (7-12). These studies demonstrated that less than half of the patients with mHNPC were treated with up-front combination therapy, and the patients mainly received ADT monotherapy or ADT plus first-generation antiandrogen (combined androgen blockade; CAB). The obvious disconnect between clinical trial evidence and real-world practice shows the unmet need for an optimal treatment strategy for patients with mCRPC, whose survival is generally <3 years (6, 13, 14). The best first-line treatment of mCRPC remains unclear because of the lack of direct comparative randomized data and well-validated practical predictive markers (4). Even a phase III, multicenter, randomized, controlled trial revealed no survival difference between enzalutamide and abiraterone in patients with CRPC before chemotherapy (15). Considering the absence of well-validated predictive biomarkers, international guidelines support the selection of any approved first-line systemic therapy (16, 17). Consequently, the choice of first-line therapy for patients with mCRPC is on the basis of clinical factors and practical considerations, such as the presence of comorbidities and drug availability.
In most patients with mHNPC who received hormone therapy, disease progression to CRPC is inevitable in spite of castrate testosterone level. Typically, CRPC is almost always associated with serum prostate-specific antigen (PSA) rise, implying that the disease continues to rely on AR signaling. In fact, randomized phase III clinical trials have shown the survival benefits of NHT that further suppresses AR signaling in patients with mCRPC (4, 18, 19). In the PREVAIL study, a double-blind trial of first-line enzalutamide showed significant survival benefits compared with placebo in patients with mCRPC (18). Moreover, the final PREVAIL analysis also revealed that greater decreases in confirmed PSA were associated with greater 5-year overall survival (OS) (20). Hence, we hypothesized that, on the basis of the AR-dependent mechanisms of NHT, the PSA kinetics in primary hormone therapy for mHNPC may be associated with the efficacy of NHT in patients with mCRPC. Of note, a multicenter study identified PSA kinetics (e.g., “nadir PSA levels” and “PSA change rate in 3 months”) during primary ADT for mHNPC as predictors for the effectiveness of NHT for mCRPC (21). We previously reported that the ratio of the initial-to-nadir PSA levels (I/N PSA) in primary ADT for mHNPC significantly predicted the response to enzalutamide and the survival in patients with mCRPC, including before or after docetaxel (22).
The current study aim was to determine if I/N PSA in mHNPC is associated with the response to first-line enzalutamide in mCRPC.
Patients and Methods
Patient population. A retrospective study was performed in accordance with the principles of the Declaration of Helsinki and approved by the Ethics Committee of the University of Occupational and Environmental Health, Japan (UOEHCRB21-074). An opt-out methodology on the institution’s website served as the mechanism for informed consent to participate, and patients who opted out were excluded from the study. The present study enrolled consecutive male patients with mCRPC at the University of Occupational and Environmental Health Hospital from August 2014 to December 2021. All study participants had histologically-proven adenocarcinoma and received primary ADT for synchronous (de novo) mHNPC following biochemical confirmation of CRPC with a serum testosterone level of ≤50 ng/dl and first-line treatment with enzalutamide. CAB consisted of castration (surgical or chemical) plus a first-generation antiandrogen (bicalutamide 80 mg/day or flutamide 375 mg/day). The exclusion criteria included metachronous/relapsing mHNPC recurring after definitive therapy given with curative intent in a localized setting, non-metastatic disease, up-front combination therapy (prior use of NHT or docetaxel for mHNPC) or ADT monotherapy (surgical or chemical castration alone), prior use of chemotherapy/NHT except enzalutamide, all of which may affect nadir PSA levels and OS (Figure 1). Patients received enzalutamide at the starting dose of 80, 120, or 160 mg/day. Dose reduction or escalation of enzalutamide between 80 and 160 mg/day were allowed for patients with tolerability, comorbidities or severe treatment-related adverse events. CRPC was defined using the Prostate Cancer Clinical Trials Working Group 3 criteria (23), while radiological assessments in routine surveillance were not conducted for all the participants. Serum PSA levels were measured every month for several months in patients with mHNPC, and every month until nadir PSA levels in patients with mCRPC during first-line enzalutamide treatment. Radiologic surveillances were performed at the discretion of each physician after enzalutamide initiation.
Patient selection flow of the present study.
The following patient background characteristics were collected: age, Eastern Cooperative Oncology Group Performance Status, primary tumor stage, International Society of Urological Pathology (ISUP)-grade group at biopsy, extent of disease score on bone scintigraphy, visceral metastases, initial PSA (iPSA) levels, nadir PSA levels, I/N PSA, time-to-nadir PSA, PSA doubling time on ADT, time to CRPC, and PSA levels at the initiation of enzalutamide. I/N PSA was calculated by dividing the iPSA level by the nadir PSA level, as previously defined (22), with minor modifications. Undetectable nadir PSA levels (<0.008 ng/ml) were treated as 0.008 ng/ml for the receiver operating characteristic (ROC) curve. Undetectable nadir PSA levels (<0.008 ng/ml) were categorized as high I/N PSA. Computed tomography and bone scintigraphy were conducted to determine the metastatic status before initiating CAB and enzalutamide. The PSA response was defined as ≥90% decline in PSA during enzalutamide administration from baseline in patients with mCRPC. PSA response rate was defined as the proportion of PSA responders among all eligible patients. OS and prostate cancer-specific survival (PCSS) were defined as the interval from enzalutamide initiation to all-cause death and PC-related death, respectively. Radiographic progression-free survival (rPFS) was defined as the interval from enzalutamide initiation to the date of progression in accordance with the criteria of the Prostate Cancer Working Group 3 (23). Patients were censored when the event did not occur within the duration of follow-up.
Treatment protocol. All patients with mHNPC received primary CAB. After mCRPC diagnosis, the patients were treated with first-line enzalutamide. Therapeutic agents for CRPC were changed following PSA and/or radiographic progression or severe treatment-related adverse events. Computed tomography and/or bone scans were conducted when patients developed PSA progression and/or any symptoms.
Statistical analysis. Fisher’s exact test for nominal variables or The Mann-Whitney U-test for continuous variables were performed to compare the patient characteristics between the groups. The optimal cutoff value for I/N PSA or nadir PSA levels was determined on the ROC curve closest to the upper-left corner. PSA response rate was used as a pretest probability for calculating the positive predictive value and the negative predictive value of I/N PSA. The patients were categorized into the low or high I/N PSA groups and nadir PSA groups according to the optimal cutoff values. The patient characteristics, rPFS, and OS were compared between the groups. The Kaplan-Meier method was performed to calculate survival rates, and the log-rank test was used for comparisons. We determined the cutoff values of non-normal distributed data using the medians for univariate and multivariate analyses. Multivariate Cox proportional hazard regression was performed to identify independent risk factors for OS. Variables with a value of p<0.05 were selected in the univariate models. All statistical analyses were performed with EZR, which is a graphical user interface for R (The R Foundation for Statistical Computing, Vienna, Austria). A value of p<0.05 was considered statistically significant.
Results
Patient characteristics. Overall, 119 patients with mCRPC received enzalutamide during the study period. 91 patients were excluded from all analyses because they were ineligible after applying the exclusion criteria or because of an absence of follow up and clinical data. The remaining 28 patients were included in the analysis.
The ROC curve analysis for PSA response demonstrated that the optimal cutoff value of I/N PSA was 1219 [sensitivity=71.4%, specificity=92.9%, area under the curve (AUC)=0.85], and that of the nadir PSA levels was 0.521 ng/ml (sensitivity=85.7%, specificity=78.6%, AUC=0.79) (Figure 2A and B). The positive predictive value and the negative predictive value of I/N PSA were 91.0% and 76.5%, respectively. Notably, there was no predictive cutoff iPSA level for PSA response (AUC=0.58).
Receiver operating characteristics curve analyses of (A) I/N PSA and (B) nadir PSA levels for predicting PSA response in patients with mCRPC. The area under the curve (AUC), sensitivity, and specificity are shown. The black circles indicate the optimal points.
The baseline patient characteristics stratified by I/N PSA are described in Table I. The median follow-up time from enzalutamide initiation for patients with mCRPC was 23.5 months (interquartile range=16-34 months). The median enzalutamide treatment duration was 10.0 months (interquartile range=4-15 months). The number of deaths after enzalutamide initiation was 19 (68%). 17 (60.7%) patients showed low I/N PSA and 11 (39.3%) patients high I/N PSA (median=70.5 and 4,912.5, respectively). The ISUP-grade group and nadir PSA levels were statistically higher in the low I/N PSA cohort than in the high I/N PSA cohort. The time-to-nadir PSA and time to CRPC were statistically lower in the low I/N PSA cohort than in the high I/N PSA cohort. Intriguingly, PSA response rate was statistically lower in the patients with low I/N PSA (23.5%) than in those with high I/N PSA (90.9%) (p=0.001). No statistically significant differences were found in the other characteristics between study groups.
Patient characteristics according to I/N PSA status.
Kaplan-Meier analyses. The median OS was 26 months [95% confidence interval (CI)=18-49 months]. A cutoff I/N PSA level of 1219 or a cutoff nadir PSA level of 0.521 ng/ml was used to investigate the heterogeneity of the PSA response. The median OS was significantly greater in patients with high I/N PSA compared with those with low I/N PSA (80 months, 95%CI=15 months–not reached vs. 20 months, 95%CI=11-29 months; p=0.001) (Figure 3A). The 3-year OS rates for patients with high and low I/N PSA were 69.3% and 14.7%, respectively. The PCSS and rPFS also showed patterns similar to those for OS (Figure 3B and C). Patients with high I/N PSA had significantly greater median PCSS (80 months, 95%CI=15 months–not reached vs. 25 months, 95%CI=11–not reached; p=0.029) and median rPFS (58 months, 95%CI=6 months–not reached vs. 11 months, 95%CI=3-15 months; p=0.014) than patients with low I/N PSA. The 1-year rPFS rates for patients with high and low I/N PSA were 87.5% and 35.2%, respectively.
Kaplan-Meier curves for overall survival (OS) (A), prostate cancer-specific survival (PCSS) (B), and radiographic progression-free survival (rPFS) (C) in patients with metastatic castration-resistant prostate cancer (mCRPC) treated with enzalutamide stratified by the optimal I/N PSA. Kaplan-Meier curves for OS (D), PCSS (E), and rPFS (F) in patients with mCRPC treated with enzalutamide stratified by the optimal nadir PSA level. The log-rank test was used to calculate the p-values.
The nadir PSA of 0.521 ng/ml demonstrated patterns that were the same as those observed for I/N PSA but with higher p values. The OS (p=0.041), PCSS (p=0.036), and rPFS (p=0.027) were significantly longer for patients with low nadir PSA levels than those with high nadir PSA levels (Figure 3D-F).
Analyses of prognostic factors. Table II demonstrates the results of the univariate and multivariate Cox regression analysis of OS after initiation of enzalutamide. Univariate analysis showed that a significant association between I/N PSA ≥1,219 and long-term survival. EOD score at the initiation of ENZ (p=0.034) was excluded from the multivariable analysis because only 14 patients (50%) were available to evaluate the factor. Multivariable analysis elucidated that I/N PSA ≥1,219 (HR=0.23, 95%CI=0.06-0.84, p=0.026) and time to CRPC ≥11 months (HR=0.32, 95%CI=0.11-0.96, p=0.041) were independent predictors of OS.
Univariate and multivariate analyses to predict overall survival.
Discussion
The present study results demonstrated that I/N PSA was associated with PSA response to enzalutamide and was an independent predicter of OS in mCRPC patients who received first-line enzalutamide. The optimal cutoff I/N PSA, nadir PSA, and initial PSA levels were analyzed, and the results revealed that the I/N PSA and nadir PSA levels were significantly associated with PSA response, rPFS, PCSS, and OS after enzalutamide initiation. Univariate analysis demonstrated that I/N PSA was significantly associated with OS, whereas nadir PSA levels and iPSA levels were not. Finally, Cox hazard model analysis revealed that I/N PSA was an independent predictor of OS with higher HR than time to CRPC in mCRPC patients. Notably, I/N PSA and time to CRPC during CAB in mHNPC might reflect the hormone sensitivity of CRPC in addition to that of hormone-naïve PC. Thus, I/N PSA can be a predictive biomarker for first-line enzalutamide in practice, which helps to select good candidates with mCRPC for first-line enzalutamide.
To the best of our knowledge, this is the first study to demonstrate a practical predictive biomarker of response to first-line enzalutamide in mCRPC. We evaluated the efficacy of I/N PSA for reasons that were previously explained (22). Briefly, PSA kinetics during primary hormone therapy, which are considered as prognosticators in patients with mHNPC, may partially reflect hormone sensitivity of mCRPC. Notably, several studies have demonstrated that low iPSA level was a poor prognostic factor reflecting poor response to ADT in patients with PC (24-26). Therefore, we combined iPSA and nadir PSA into I/N PSA, which may reflect the response to further suppression of AR signaling in mCRPC. Our study results showed that I/N PSA was a predictive and prognostic biomarker for first-line enzalutamide in patients with mCRPC.
The ability of I/N PSA to predict the efficacy of enzalutamide may be explained by tumor heterogeneity in terms of AR activation. Nelson described four discrete molecular states of PC defined on the basis of the androgens/androgenic ligand sources and AR activity (27). According to the molecular states, enzalutamide may be effective against States 1-3 but not State 4: androgen- and AR-independent neuroendocrine PC, which may show low iPSA and poor response to ADT (25). Additionally, the molecular states and our results allow for a hypothesis that mHNPC with higher I/N PSA during primary CAB may develop a higher rate of State 2-3 clones, whereas mHNPC with low I/N PSA, particularly with low initial-high nadir PSA levels, may have homogeneous clones, including State 4 clones. Our results do not disprove that hypothesis.
Remarkably, three real-world retrospective studies consistently revealed underutilization of up-front therapy in patients with mHNPC despite level 1 evidence and guideline recommendations (10-12). Using the Veterans Health Administration claims database from 2013 to 2018, Freedland et al. (10) found that 87% of male patients with mHNPC received ADT monotherapy or CAB as first-line therapy. Even in the period from 2017 to 2018, only 24% of the patients were treated with up-front therapy using docetaxel or NHT. Ryan et al. (11) reported that only 13% of patients were treated with docetaxel or abiraterone acetate in 2019. Finally, using the IQVIA open-claims database that included information from approximately 18 million U.S. cancer patients, Heath et al. (12) described the first-line treatment patterns observed in mHNPC patients. Surprisingly, in 2021, 64% of patients with mHNPC still received ADT monotherapy or a first-generation antiandrogen as first-line therapy. These real-world data clearly demonstrate unmet needs for optimizing first-line treatment in patients with mCRPC after primary ADT even in the up-front era.
There are several limitations to this study. First, the study was retrospective non-randomized with a relatively small cohort and the results might be confounded by unobserved factors. Second, the results only demonstrated the value of I/N PSA as a predictive biomarker for first-line enzalutamide in patients with mHNPC received primary CAB, but not with ADT monotherapy. Since nadir PSA levels were higher after initiation of ADT monotherapy than after CAB (28), the cutoff value for predicting response to enzalutamide may be lower for I/N PSA in ADT monotherapy than for CAB. Third, the results also demonstrated that the value of “I/N PSA” or “PSA decline” as a predictive biomarker only for first-line enzalutamide, but not for other NHT, such as abiraterone (22, 29). Moreover, the cutoff value of 1,219 in this training cohort needs to be investigated in a validation cohort. Large-scale, prospective studies are warranted to further validate our results.
In conclusion, these results demonstrated that I/N PSA predicted response to first-line enzalutamide in patients with mCRPC, a characteristic that could help optimize the therapeutic strategies for the population. Moreover, I/N PSA was found to be a prognosticator in these patients. Further large-scale and prospective studies are warranted.
Acknowledgements
The Authors would like to thank Enago (www.enago.jp) for the English language editing.
Footnotes
Authors’ Contributions
YN: Project development, data curation, data analysis, manuscript writing and editing. KJ: data curation and critical review. TM: data curation and critical review. IT: supervision. NF: project development, writing, and supervision.
Conflicts of Interest
The Authors declare that there was no financial support or conflicts of interest related to this study.
- Received July 21, 2023.
- Revision received August 26, 2023.
- Accepted August 28, 2023.
- Copyright © 2023 The Author(s). Published by the International Institute of Anticancer Research.
This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY-NC-ND) 4.0 international license (https://creativecommons.org/licenses/by-nc-nd/4.0).
