Introduction

Prostate cancer (PCa) is the most common malignancy in men throughout the world [1]. Active surveillance, radical prostatectomy (RP) and radiotherapy (RT) are the main definitive treatment methods of the disease for localized stages [2, 3]. RT can also be used postoperatively as an adjuvant or salvage treatment. More than 60% of men with adverse pathologic risk factors such as extraprostatic extension (EPE), seminal vesicle involvement (SVI) and positive surgical margins develop a biochemical recurrence (bR) within 10 years after RP. Former randomized trials concluded that adjuvant RT halves bR in patients with these adverse risk factors [4,5,6]. More recently, RADICALS-RT, RAVES and GETUG-17 trials have found that early salvage RT ± androgen deprivation therapy (ADT) has disease control rates comparable to adjuvant RT with low urinary morbidity [7,8,9]. The pre-planned ARTISTIC collaborative meta-analysis of these three randomized trials also confirmed these results [10]. In addition to prostate bed RT, the benefit of ADT in the salvage setting has been shown in RTOG 9601 and GETUG-AFU 16 [11, 12]. Moreover, NRG Oncology/RTOG 0534 SPPORT reported that the combination of ADT and pelvic RT adjunct to prostate bed RT provided better freedom from progression rates [13]. In the era of ultra-sensitive prostate-specific antigen (PSA) analysis, early salvage RT is widely accepted by clinicians in low-intermediate risk patients. However, there are still controversial points regarding the timing of postoperative management of PCa after RP in high-risk patients. Besides, there is no consensus on the use of ADT and pelvic RT in the postoperative setting. The definition of the patient at high risk and the definition of recurrence; timing, dose, and target volume of RT; the addition, duration and form of ADT are the main research topics of the latest randomized trials.

In daily practice, clinicians have different approaches to patients with adverse risk factors. Some clinicians consider certain risk factors, while others consider the combination of these when deciding on the treatment scheme. In this study, we aimed to analyse the salvage treatment approaches and treatment outcomes for relapsed PCa patients treated in four centers retrospectively.

Material and method

This study is a multicenter, retrospective evaluation of salvage RT outcomes of the PCa patients treated at four radiotherapy centers between 2007 and 2021. A total of 300 consecutive patient data were collected from 2 public, 1 academic and 1 private hospital. Eligibility criteria after radical prostatectomy were as follows; 1. Persistently detectable PSA level, 2. PSA ≥ 0.1 ng/ml at any time, 3. Three consecutive PSA rises after an initially undetectable level. Patients who underwent adjuvant RT at undetectable PSA levels, who were followed earlier than a year, and who had a pre-existing malignancy were excluded from the study. There were 272 patients included in the final analysis. TNM staging was performed according to the 8th edition of the American Joint Committee on Cancer (AJCC) staging system. The Common Terminology Criteria for Adverse Events v4.0 was used to evaluate acute and late genitourinary and gastrointestinal side effects. The study protocol was approved by the Local Ethics Committee (91/04.04.2022).

Statistical analysis

Categorical variables were presented with frequency and proportion, and continuous variables with median, minimum and maximum values. A Receiver Operating Characteristics (ROC) analysis was used to find the predictive cut-off value of pre-RT PSA for disease progression. Biochemical relapse-free survival (bRFS) was calculated from the end of the salvage RT to the second PSA relapse. PSA relapse after salvage RT was defined as PSA level ≥ 0.4 ng/mL or second initiation of salvage therapy. Clinical relapse-free survival (cRFS) was calculated from the end of salvage RT to the time of lymph node, bone, or visceral metastasis. Overall survival (OS) was calculated from the end of salvage RT to any cause of death. Kaplan–Meier method was used to calculate the 5-years bRFS, cRFS, and OS rates. Univariate analyses of time to bR and clinical relapse were conducted using Kaplan–Meier plots and log-rank tests. Multivariate analyses of time to bR and clinical relapse were performed using a Cox proportional hazards model with 95% confidence intervals (CIs). Statistical analyses were performed with SPSS version 20 for Windows (IBM Corp., Armonk, NY). The p value was set at < 0.05 for significance.

Results

The median age was 65 (48–82) years. Staging before surgery was performed with Magnetic resonance imaging (MRI) in 128 (47%) patients, bone scintigraphy in 10 (3.7%) patients and 68Ga-PSMA-PET/CT in 55 (20.2%) patients. All patients had radical prostatectomy; laparoscopic surgery was used in 71 (26.1%), and pelvic lymph node dissection was performed in 155 (57%) patients. A total of 42 (15.4%) patients had pathologically involved nodes. Patients’ and disease characteristics are given in Table 1.

Table 1 Patients’, disease and treatment characteristic
Full size table

Treatment after the first PSA relapse

Persistent PSA (≥ 0.1 ng/ml) after RP was present in 132 (48.5%) of the patients. MRI was used for restaging at relapse in 118 (43.4%) patients and 68Ga-PSMA-PET/CT was performed additionally in 71 (26.2%) patients. Of these, 26 (9.6%) had local, 10 (3.7%) had pelvic lymphatic and 4 (1.5%) had local + pelvic lymphatic recurrence. Salvage RT was implemented a median of 8 (4–149) months after RP. RT was delivered mostly with intensity-modulated radiation therapy (IMRT) in 232 (85.3%) patients and with three-dimensional conformal radiotherapy (3D-CRT) in 40 (14.7%) patients. RT target volumes consisted of prostate bed alone in 206 (75.7%) and prostate bed + pelvic lymph nodes in 66 (24.3%) patients. The median dose to prostate bed and pelvic lymphatics were 66 (60–74) Gy and 50 (45–56) Gy, respectively. Thirteen patients received lymphatic boost up to a median of 66 (60–78) Gy in addition to pelvic lymphatic irradiation. ADT was added in 158 (58.1%) patients. Treatment characteristics are given in Table 1.

Side effects

Overall, side effects were generally mild and RT was well tolerated. Grade 3–4 acute genitourinary and gastrointestinal side effects were seen in only 1 patient. Grade 3–4 late genitourinary and gastrointestinal side effects were seen in 6 (2.2%) and 4 (1.5%) patients, respectively.

Follow-up

The median follow-up time was 64 (12–180) months. Each patient was followed-up according to the protocols of each institute. The patients were followed-up every 3 months for the first 2 years, every 6 months till 5th year, and annually thereafter. PSA was evaluated in addition to a physical examination at each visit. Imaging interventions were performed in case of a rising PSA.

Treatment outcomes

Biochemical relapse after salvage treatment was observed in 49 (18%) patients and clinical relapse was observed in 34 (12.5%) patients. Clinical relapse was seen in the pelvic lymphatics in 11, in the paraaortic lymphatics in 4, in the bone in 17, and in the visceral organs in 2 patients. At the time of evaluation, 27 (9.9%) patients died, and only 4 (1.5%) died due to prostate cancer. 5-years bRFS, cRFS, and OS were found 75.1%, 84.8%, and 94.9% respectively. The cut-off PSA value for lower bRFS was found to be 0.14 (AUC 0.607, p 0.019) in ROC analysis.

In univariate analysis, 5-years bRFS was found to be lower for patients ≤ 65 years old than for patients > 65 years old (68.4%, 83.8%, p 0.006), patients with pre-RT PSA level > 0.14 ng/mL than patients with ≤ 0.14 ng/mL (71.2%, 89.2%, p 0.035, Fig. 1a), patients with stages pT3b vs pT3a vs pT2 (49.1%, 82.8%, 90.3%, p < 0.001, Fig. 1b), and patients with pathologic pelvic lymph node involvement ≥ 2 vs 1 vs 0 (38.5%, 66.8%, 93.7%, p < 0.001). A trend for lower bRFS was also observed with increasing Gleason scores (p = 0.054). We did not find any difference between patients treated with and without pelvic lymph node RT, or between patients treated with and without ADT. This may be due to treatment selection bias. Not surprisingly, the proportion of patients with pN1, GS ≥ 8, and pT3b was higher in patients treated with pelvic lymph node RT and/or ADT. In multivariate analyses, stages pT3a-b, pre-RT PSA value above 0.14 ng/ml, and two or more pelvic lymphatic involvement were found independent adverse prognostic factors for bRFS. SVI was found to be the most significant adverse risk factor for disease progression (HR 8.6 95% CI, 3.4–21.5, p < 0.001). Results of univariate and multivariate analyses of bRFS are listed in Table 2. Similar results remained valid in the cRFS analysis.

Fig. 1
figure 1

Kaplan–Meier plots showing freedom from biochemical relapse by pre-radiotherapy PSA level (a) and pT stage (b)

Full size image
Table 2 Univariate and multivariate analysis of bRFS
Full size table

Dıscussion

The optimal management of biochemically relapsed PCa after RP remains unclear. Several patient and disease characteristics are considered for the best-individualized salvage management strategy for PCa. In this study, 5-year bRFS for the whole cohort was 75.1% and the most important independent factors for worse disease control were found to be SVI, ≥ 2 pathologic nodal involvement and higher PSA levels (> 0.14 ng/ml) during salvage treatments. These findings are consistent with the latest randomized trials.

In NRG Oncology/RTOG 0534 SPPORT Trial; SVI, Gleason score of 8–9, and pre-RT PSA > 1 ng/mL were found to be significant adverse factors of freedom from progression. 5-years progression-free survival rates were improved from 70.9 to 81.3% with short-term ADT and to 87.4% both with short-term ADT and pelvic RT in addition to prostate bed RT. The 8-years estimated freedom from progression rate was worst in patients with SVI (37.2%), and improved to 50.8% when short-term ADT was added and to 59.6% when short-term ADT plus pelvic RT added [13]. In GETUG-AFU 16 study, better progression-free survival was achieved with 6 months of goserelin (5-years bRFS, 80% vs 62%, p < 0.0001). Similar to NRG Oncology/RTOG 0534 SPPORT, pre-RT PSA level and SVI were found to be unfovarable factors for relapse [12]. The advantage of long-term ADT in the salvage setting has also been demonstrated in two trials. In RTOG 9601 study, the 12-years incidence of prostate cancer mortality was 5.8% with 2 years of bicalutamide versus 13.4% with placebo [11]. In recently reported first results of the RADICALS-HD trial, 24 month ADT improved metastasis-free survival rates than 6 months ADT (HR 0.77; CI 0.61–0.97; 72% vs 78%) at 10 years [14]. In our study, 42% of the whole cohort and 30% of the patients with SVI did not use ADT. Adding at least 6 months of ADT could improve the bRFS rates of patients with SVI as well as the entire cohort. In the same way, pelvic RT was implemented in 24.3% of our patients. In our study staging with 68Ga-PSMA-PET/CT before salvage treatments was done in only 71 (26%) of patients. Pelvic lymph nodes were involved in 14 (20%) of them. These patients underwent pelvic RT with a boost dose to the involved lymph node(s) and no bR was observed until the time of evaluation. This underlines that 68Ga-PSMA-PET/CT must be performed prior to salvage RT for re-staging and modification of RT field and dose.

Early implementation of salvage RT is also important due to concerns about the progress to the metastatic stage. Tendulkar et al. reported that in a large, multicenter dataset of 2460 patients undergoing salvage RT for PCa, 5-year bRFS rates were 71% when salvage RT was performed at PSA levels of 0.01–0.2 ng/mL and decreased to 37% when PSA level was > 2.0 ng/mL. In this study, the overall 5-year bRFS was lower than our series (56% vs. 75%). This may be due to higher pre-RT PSA levels (median 0.50 ng/mL vs. 0.35 ng/mL) and a lower rate of ADT use (16% vs. 58%) in their patient cohort [15]. In the RADICALS-RT, RAVES and GETUG-AFU 17 trials, the median PSA before salvage RT was around 0.20–0.24 ng/mL and compared with historical data, the rates of disease control was higher than 85%. However, these randomized clinical trials have been criticized for the small number of high-risk patients (pN1, pT3b and GS ≥ 8), which may underpower subgroup analysis even at longer follow-ups. A huge study cohort of 26.118 patients recently reported by Tilki et al. found that adjuvant RT reduced all-cause mortality among men with pN1, GS ≥ 8 and pT3/T4 stages (HR 0.33, 0.13–0.85, p 0.02) [16]. They support adjuvant RT instead of early salvage RT in this high-risk patient group in their commantery [17]. Abdollah et al. also demonstrated the OS advantage of adjuvant RT in the presence of at least two of these risk factors [18].

Recent AUA/ASTRO guidelines state that patients with adverse risk factors, including R1, EPE, and SVI, “should be offered adjuvant RT”. EAU/ESTRO guidelines recommend adjuvant RT for patients with two of the risk factors, R1, pT3, and Gleason score ≥ 8. Both underlines the initiation of salvage RT at the earliest sign of PSA rise. ADT was strongly recommended in the AUA/ASTRO guidelines and weakly in the EAU/ESTRO guidelines for salvage setting, however, the duration was not pronounced [2, 19]. High-risk patients with SVI, GS ≥ 8, and positive pelvic nodes needs special attention. Early salvage prostate bed + pelvic RT and long-term ADT may be considered in patients with SVI and ≥ 2 positive pelvic lymph nodes. Given the high rates of comorbities in this age group, Adjuvant RT may also be an option in some of these patients who can not tolerate ADT and/or pelvic RT.

The main limitation of our study was its retrospective nature. Treatment biases based on risk factors limit the interpretation of results. Patients included in the study were treated at four different centers over a 15-years time period. The treatment protocols can be distinct and can have alterations during the years. 68Ga-PSMA-PET/CT was used for staging in a small number of patients both before the surgery and salvage RT. Some metastatic patients may have been undetected. PSA doubling time, which is an important factor, was not reached in all patients in this study. It can facilitate the postoperative management of PCa. Finally, the 5-years follow-up period may not be sufficient for PCa and longer follow-up may be required. However, the study addresses the real-life data of salvage treatment options for PCa.

Conclusion

Salvage RT ± ADT provided 5-year biochemical disease control in 75.1% patients. SVI, ≥ 2 positive pelvic nodes, and delayed administration of salvage RT (PSA > 0.14) were found to be independent adverse risk factors for the second relapse after salvage RT for PCa. Such factors should be taken into account during the decision process on salvage treatment.