Analysis of risk factors for disease progression after salvage radiation therapy with androgen deprivation therapy in prostate cancer patients who have prostate-specific antigen persistence after radical prostatectomy
Article information
Abstract
Purpose
To assess risk factors of disease progression after salvage radiation therapy (SRT) with androgen deprivation therapy (ADT) in case of prostate-specific antigen (PSA) persistence after radical prostatectomy (RP).
Materials and Methods
We analyzed 57 patients who received SRT with ADT between 2013 and 2019 due to PSA persistence after RP. The endpoint was disease progression defined by biochemical recurrence or clinical recurrence. Age, Pre-RP PSA level, Gleason score, pathologic stage, presence of pelvic lymph node dissection, surgical margins, and PSA at 6-8 weeks after RP were analyzed as predictive factors for disease progression. Kaplan-Meier method and Cox regression models were used for data analysis.
Results
At a median follow-up of 38 months (interquartile range, 26–61), 17 patients had disease progression. Pathologic T stage (pT3b vs. pT3a or lower; hazard ratio [HR] = 9.20; p = 0.035) and PSA level at 6-8 weeks after RP (≥2.04 vs. <2.04 ng/mL; HR = 5.85; p = 0.002) were predictors of disease progression. The 5-year disease progression-free survival rate was 46.7% in pT3b group as compared to 92.9 % in pT3a or lower group, and 18.4% for PSA ≥2.04 ng/mL after RP as compared to 79.2% for PSA <2.04 ng/mL.
Conclusion
Pathological T stage (pT3b) and post RP PSA ≥2.04 ng/mL are independent risk factors of disease progression after SRT with ADT in patients with PSA persistence after RP.
Introduction
Radical prostatectomy (RP) is one of the standard treatment modalities for patients diagnosed with localized and locally advanced prostate cancer (PCa). Due to the half-life of prostate-specific antigen (PSA), serum PSA levels are expected to be nondetectable after 6 weeks from RP [1]. However, up to 20% of post-RP patients experience persistently detectable PSA level and this is a predictor of both disease progression and worse cancer-specific survival [2-4]. In the ARO 96-02 phase-III trial on salvage radiation therapy (SRT) after RP, patients with PSA persistence after RP showed worse 10-year metastasis-free survival (MFS; 67% vs. 83%) and overall survival (OS; 68% vs. 84%) than patients who had post-RP undetectable PSA [5]. And the post-hoc analysis of Radiation Therapy Oncology Group 9601 noted that patients with PSA persistence after RP showed worse local and metastatic disease occurrence rates and overall-mortality rates compared to patients with biochemical recurrence (BCR) after RP [6].
SRT and androgen deprivation therapy (ADT) are recommended therapeutic options for high risk PCa patients with BCR after RP and resulted in significantly higher long-term overall survival and lower metastasis rates [7-9]. Nevertheless, few studies have assessed the oncological outcomes of SRT with ADT in patients with post-RP PSA persistence, and salvage management remain controversial. Identifying predictors of disease progression may result in improvement of the treatment and follow-up strategies.
In this retrospective study we examined the oncologic and survival outcomes and predictors of disease progression in patients treated with SRT and ADT for persistently detectable PSA after RP.
Materials and Methods
1. Patients
This study was approved by the Pusan National University Hospital Ethics Committee (Approval No. 2019-005-106). We retrospectively reviewed histologically confirmed pT2–3 PCa patients who have undergone SRT and ADT after RP with or without pelvic lymph node dissection between 2013 and 2019. Inclusion criteria were as follows: RP for localized PCa, persistently elevated PSA as defined by PSA ≥0.1 ng/mL at 6–8 weeks after RP, available postoperative PSA assay. During the study period, a total of 119 patients received radiotherapy (RT) after prostatectomy, only one patient received adjuvant RT and all others received SRT. Among the patients who received SRT, 62 patients showed PSA persistence. Of the patients with PSA persistence, 57 patients were analyzed except for one who did not receive ADT and four due to loss of follow-up.
2. Treatment
The median time between RP and SRT was 7 months (interquartile range [IQR], 5–13). SRT was performed after a certain period of time after RP for recovery of urinary continence. The median dose of SRT was 70 Gy (IQR, 66–70) using conventional fractionation (2 Gy/fraction) intensity-modulated radiation therapy technique. The prostate and seminal vesicle bed were treated in all patients. Thirty-two patients received additional pelvic RT. The median time from RP to ADT was 4.4 months, and 51 of 57 patients received ADT before SRT. The median duration of ADT before disease progression was 21 months (IQR, 11–28). Fifty patients were treated with gonadotropin-releasing hormone (GnRH) agonist and antiandrogen, and seven patients were treated with only GnRH agonist.
3. Outcomes
The disease progression was defined as BCR or clinical recurrence after SRT with ADT. The BCR was defined as two consecutive serum PSA level >0.2 ng/mL. Clinical recurrence was defined as local recurrence and distant metastasis confirmed by radiologic imaging consisting of computed tomography (CT), magnetic resonance image (MRI), and bone scan. BCR, clinical recurrence, and death from any cause was the endpoint of disease progression-free survival (DFS).
4. Statistical analysis
Age, Pre-RP PSA level, Gleason score, pathologic T stage, N stage, presence of pelvic lymph node dissection (PLND), surgical margins, pelvic lymph node radiation, and PSA at 6–8 weeks after RP were analyzed as predictors for disease progression after SRT with ADT. The pathologic T stage was classified as ≤pT3a or >pT3b, and GS was classified as ≤8 or ≥9. Pathologic T stage and Gleason score were divided into two groups based on the results showing the highest significance level in the univariate analysis. In analyzing N stage, the pathologic stage was used for patients who underwent PLND, and the clinical stage was used for patients who did not undergo PLND. Follow-up period was defined as the time from completion of SRT to disease progression or last follow-up. Kaplan-Meier curve analysis was used to estimate DFS. Cox regression analysis using hazard ratio (HR) with their 95% confidence intervals (CI) was performed to evaluate predictive factors. Statistical analyses were conducted using SPSS software version 22.0 (IBM, Armonk, NY, USA) with a two-sided significance level of p < 0.05.
Results
Table 1 shows the patients' characteristics. Median age at the start of SRT was 67 years (IQR, 63–70). Gleason score was ≥9 in 35.1% of cases. pT3b stage was reported in 63.2% of cases. The median pre-RP PSA and PSA at 6–8 weeks after RP were 25.76 ng/mL (IQR, 14.03–49.76) and 0.75 ng/mL (IQR, 0.26–2.82).
Patient and tumor characteristics
At a median follow-up of 38 months (IQR 26–61), 17 patients had disease progression. Of those 17 patients, all patients presented biochemical recurrence, 6 (35.3%) had clinical recurrence including 1 (5.9%) with local recurrence, 5 (29.4%) with distant metastasis, and 2 (11.8%) deaths during the follow-up period. In all patients who had clinical recurrence, BCR occurred before clinical recurrence. The specific site of metastasis consisted of bone (n = 4), pelvic lymph nodes (n = 1), retroperitoneal lymph nodes (n = 1), and neck lymph nodes (n = 1). The mean PSA at 6–8 weeks after RP of patients with BCR was 2.8 ng/mL (1.8 ng/mL for all patients), and the pT3b rate of those was 94.1% (63.2% for all patients).
Table 2 presents the results of univariate Cox regression analysis. The following factors showed significant association with disease progression: PSA at 6–8 weeks after RP (p < 0.001), pathologic T stage (p = 0.013). Fig. 1 shows the receiver operating characteristics (ROC) curve for the prediction of disease progression according to PSA level at 6–8 weeks after RP. Optimal cutoff was PSA of 2.04 ng/mL (area under the curve [AUC] = 0.682; 95% CI 0.526–0.837; p = 0.031). Pre-RP PSA, Gleason score, and surgical margins were not prognostic. Optimal cutoff point analyzed by ROC curve of pre-RP PSA was 20.9 ng/mL (AUC = 0.649; 95% CI 0.505–0.792; p = 0.078). On multivariate analysis considering PSA at 6–8 weeks after RP and pathologic T stage, both PSA at 6–8 weeks after RP (HR = 5.85; 95% CI 1.88–18.20) and pathologic T stage (HR = 9.20; 95% CI 1.17–72.51) were independent predictors of increased risk of disease progression (Table 3).
Univariate Cox regression analysis for disease progression
Receiver operating characteristics curve for the prediction of disease progression according to the prostate-specific antigen level at 6–8 weeks after radical prostatectomy.
Multivariate Cox regression analysis for disease progression
The DFS curve is shown in Fig. 2. The median DFS after SRT was 79.8 months and the 5-year DFS rate of overall population was 63.8%. Figs. 3 and 4 illustrate DFS curves stratified by the pathologic T stage and by the PSA at 6–8 weeks after RP. The 5-year DFS was 46.7% in pT3b group as compared to 92.9% in pT3a or lower group, and 18.4% for PSA ≥2.04 ng/mL after RP as compared to 79.2% for PSA <2.04 ng/mL.
Disease progression-free survival (DFS) curve of overall population.
Disease progression-free survival (DFS) curves stratified by the pathologic T stage.
Disease progression-free survival (DFS) curves stratified by the prostate-specific antigen (PSA) at 6–8 weeks after radical prostatectomy.
Discussion and Conclusion
In this study of 57 patients, 17 (29.8%) patients had disease progression and the independent risk factors of disease progression were pathologic T stage (T3b) and post-RP PSA ≥2.04 ng/mL. Post-RP PSA persistence is strongly and independently associated with adverse pathologic features and poorer survival outcomes compared with undetectable PSA and is associated with the need for salvage treatment, including SRT and/or ADT [2-4]. However, few previous studies assessed the impact of PSA persistence on oncological outcomes of patients who received SRT with ADT after RP. To address this void, we investigated predictors of disease progression, and oncologic and survival outcomes in patients treated with SRT and ADT for persistently detectable PSA after RP.
Serum PSA level is one of the most important PCa risk stratification factors. Previous studies have assessed PSA persistence itself without analysis based on PSA values. If the initial post-RP PSA values indicate the residual tumor burden, the prognosis may depend on those values. In this study, only post-RP PSA level showed significant association with disease progression and pre-RP did not. Our result was consistent with this assumption highlighting the importance of post-RP PSA value. Accordingly, the first PSA after RP should be carefully monitored. Seminal vesicle invasion by PCa has been a key prognostic factor [10]. In our study, the effect of pT3b was also observed in patients with PSA persistence after SRT and ADT.
In many studies, surgical margin was a prognostic factor for PSA persistence after RP [3,11]. In our study, we analyzed BCR and clinical recurrence after salvage treatments of RT with ADT in PSA persistent patients, and positive margin was not a risk factor for recurrence. It can be assumed that the effect of positive margin on recurrence was well controlled by salvage treatments.
In a retrospective study by Ploussard et al. [12], which assessed oncologic outcomes after SRT without ADT in patients with PSA persistence after RP, the pathological predictors of SRT failure included pT3b stage, Gleason score 8 or more, pre-RP PSA (≥20 ng/mL), PSA at 6-week after RP (≥0.4 ng/mL), pre-SRT PSA level (>1 ng/mL), and high pre-SRT PSA velocity (>0.04 ng/mL/mo). In this study, we found that PSA at 6–8 weeks after RP and pT3b were prognostic factors of DFS, which is consistent with the previous study. However, we could not analyze PSA kinetics such as PSA velocity and PSA doubling time because patients presented with short period between RP and salvage treatments.
The recent standard treatment for PSA persistence can be considered a combination of SRT and ADT [7-9]. However, there is no study yet that has analyzed the prognostic factors of patients who received SRT and ADT together as salvage treatments after PSA persistence. We think the strength of this study is that it analyzed these factors for the first time.
Our study has several limitations. First, it had a relatively small number of patients and retrospective design. This may have resulted in selection bias. The small number of patients limits the interpretation of multivariate analysis results. And as a retrospective study, there were no unified criteria for elective pelvic lymph node irradiation and when to implement SRT. Second, follow-up period was too short to analyze long-term survival rates. Third, we only used conventional imaging modalities including bone scan, CT, and MRI to detect the residual lesions after RP. Prostate-specific membrane antigen positron emission tomography (PET) is the most sensitive test for patients with PSA level <1.0 ng/mL [13]. Finally, the period and the type of ADT were not included in this analysis due to heterogeneity of ADT application. A recent randomized phase II trial (GETUG 22) which compare RT ±6 months ADT for patients with persistent PSA after RP (0.2–2 ng/mL) reported that post-RP PSA level (<0.6 ng/mL) and pathologic stage (≤pT3a) were associated with BCR-free survival. And immediate SRT with ADT improved MFS without impaired quality-of-life [14]. This is consistent with our prognostic factors.
We have confirmed that in patients with persistent PSA after RP, higher pathological T stage (pT3b) and post RP PSA ≥2.04 ng/mL are the independent risk factors of disease progression after SRT with ADT. More aggressive treatment strategies such as higher dose of SRT, pelvic lymph node irradiation, long-term ADT, and PET scans before salvage treatment should be considered for these patients.
Notes
Statement of Ethics
This study was approved by the Pusan National University Hospital Ethics Committee (Approval No. 2019-005-106).
Conflict of Interest
No potential conflict of interest relevant to this article was reported.
Funding
This work was supported by clinical research grant from Pusan National University Hospital in 2022.
Author Contributions
Conceptualization, DHK; Investigation and methodology, KHL, DHL; Writing of the original draft, KHL, DHL; Writing of the review and editing, WTK, JHN, DP, YKK, JHJ, HSJ; Formal analysis, KHL, DHK; Data curation, KHL, DCK.
Data Availability Statement
The data that support the findings of this study are available from the corresponding author upon reasonable request.