Radical prostatectomy (RP) and radiation therapy (RT) are the two first-line therapeutic options for patients with prostate cancer, with best results achieved in patients with organ-confined disease. Recurrence after RP is associated with a high Gleason score, a high level of prostate-specific antigen (PSA) before surgery, advanced tumor stage, infiltration of the seminal vesicles, or positive surgical margins [13, 35, 42, 56]. However, even in patients with favorable prognostic factors, biochemical recurrence is a common event.

Following RP, PSA should become undetectable within 4–6 weeks [52]. Persistent serum PSA levels after RP indicate residual prostate tissue, either malignant or benign. In the former case, the marker precedes clinical evidence and correlates well with disease progression.

A PSA increase above 0.2 ng/ml, a common definition of progression of disease following RP [23], occurs in up to 50% of patients with pT3/4 tumors and up to 70% of those with pT3 tumors with positive surgical margins and/or positive pelvic lymph nodes [39, 55]. The 7-year rate of biochemical progression in patients with organ-confined tumors (pT2) and positive surgical margins is about 25% [55]. Residual tumor tissue was bioptically confirmed in 35–55% of patients with rising PSA after RP without clinical correlates [46].

Here, we compare post-PR treatment approaches for high-risk patients: Adjuvant radiation therapy (ART) for men with an undetectable post-RP PSA vs. observation and salvage radiation therapy (SRT) for men with postoperatively persisting PSA or values re-rising from initial nondetectability.

Clinical target volume

The most common sites of biopsy-proven local relapse are the vesicourethral anastomosis (VUA, 66%), followed by the bladder neck (16%) and the retrotrigone area (13%) [16]. Based on the relapse patterns as determined by magnetic resonance imaging, a cylindrical shape for the clinical target volume (CTV), centered 5 mm posterior and 3 mm inferior to the VUA, has been recommended [29].

To address uncertainties in CTV definition, the Radiation Therapy Oncology Group (RTOG) [28], the EORTC Radiation Oncology Group [37], and other cooperative groups [66] have created consensus guidelines for delineation of target volumes for postprostatectomy patients. The RTOG results are available as a CT image atlas under http://www.RTOG.org. They allow for a uniform CTV definition for clinical trials that include postprostatectomy RT.

Adjuvant radiotherapy

Three randomized trials demonstrated an almost 20% benefit for biochemical progression-free survival (bNED) after ART (60–64 Gy) compared with a “wait and see” policy, mostly for pT3 cN0 or pN0 tumors (Tab. 1). The greatest advantage (30% bNED after 5 years) was seen in patients with pT3 tumors and positive margins [5, 57, 62, 64]. In fact, after central pathological review of the EORTC data, the benefit was exclusively confirmed in patients with positive margins [6, 62]. In the trial of the German Cancer Society, bNED was improved after ART: 72% vs. 54% (p < 0.03). In the subgroup of pT3 R1 tumors, bNED was still 28% vs. 18% [64]. In the study of the South Western Oncology Group (SWOG), overall survival improved from 13.5 years without ART to 15.2 years with ART [58]. The seemingly poorer results of the German trial are probably due to the reduced PSA detection threshold (Tab. 1).

Tab. 1 Overview of all three randomized trials for adjuvant radiation therapy after radical prostatectomy
Full size table
Tab. 2 Results for salvage radiotherapy after biochemical recurrence from selected studies
Full size table
Tab. 3 Recommended procedures for patients with undetectable PSA after radical prostatectomy [23]
Full size table

The location and extent of positive surgical margins after RPy are significant risk parameters of biochemical progression after RP [5]. According to the above-mentioned trials, patients with positive margins and pT3 tumors have the largest benefit from postoperative RT.

Among EORTC patients with pT2 tumors and positive surgical margins, there was a significant benefit of 5-year bNED in the irradiated group (76.4% vs. 52.2% in the wait-and-see group) [5]. However, biochemical progression was not a primary endpoint of this study. Therefore, the results must be interpreted cautiously. The benefit of radiotherapy is restricted by potential late effects such as erectile dysfunction.

Pelvic lymph nodes

The effect of ART in node-positive prostate cancer has not yet been prospectively assessed. One retrospective study reported a significant positive impact of RT combined with hormone therapy (HT) in pre-RP node-positive patients [17]. ART patients in this study were those affected by more aggressive disease. However, including ART in the multivariable models of bNED and cancer-specific survival improved the predictive accuracy significantly.

A retrospective study of 703 matched node-positive patients compared post-RP adjuvant HT plus ART with HT alone. Better survival rates were associated with combined ART/HT when patients were stratified according to nodal invasion (≤ 2 vs. > 2 positive nodes). The overall survival advantage was 19% in favor of HT plus ART [8]. Without standardized target volumes, radiation dose, and duration of HT, these data should be interpreted cautiously. However, this treatment may be indicated in selected cases, and should be validated in prospective clinical trials.

In a study of 160 patients who underwent ART or SRT, 114 were considered at high risk of lymph node involvement despite cN0 classification. Seventy-two had whole pelvic radiation and 42 had prostate bed radiation therapy. The bNED after whole pelvic radiation therapy was higher (5-year rate 47% vs. 21%, p < 0.05). While these data have to be confirmed in a prospective trial, whole pelvic radiation preferably with IMRT can be considered in high-risk patients [23, 51].

Hormone therapy and adjuvant radiotherapy

The standard nonoperative management of patients with locally advanced prostate adenocarcinoma includes long-term androgen deprivation therapy (ADT) [4, 24]. By contrast, for men who had RP and pelvic lymph node dissection for high-risk, node-negative prostate adenocarcinoma, the benefit from adjuvant ADT was not clearly established. The primary rationales for post-RP ADT is to eradicate potentially radioresistant cells in a hypoxic scar, address micrometastatic disease, and delay PSA progress in patients who will eventually relapse [22, 25, 41].

In the ongoing EORTC trial 22043, in patients with a Gleason score of 5–10, undetectable PSA, and pathological stage pT2R1 or pT3a-b, the primary trial endpoint is 5-year bNED.

Salvage radiotherapy

The best treatment in patients with positive PSA but no clinical evidence of disease is still controversial. However, only RT offers the chance of cure to patients with real local relapse. There are indicators suggesting purely local recurrence, such as a PSA doubling time of 12 months or more, more than 1-year latency in a positive post-PR PSA, a Gleason score under  7, and negative surgical margins [36]. On the other hand, a PSA doubling time of less than  12 months or a Gleason score of 8–10 at RP [33] hint instead at metastatic disease. Recently, a predictive model for the outcome of SRT after post-RP PSA progression has been established [53]. Assuming local disease, SRT of the prostate bed has been widely used to treat patients in the absence of biopsy-proven local recurrence. An established standard is conformal radiotherapy with 66 Gy to lower the risk of a ‘‘second wave of metastasis’’ [15, 23].

Diagnostic tools

Bioptic sensitivity for post-RP local recurrence ranges from 29–50% when based on digital rectal examination. With transrectal ultrasound (TRUS) guiding, it increases to 66–77%, depending on the PSA level [18, 44, 46], and with MRI even 87% can be achieved [45]. Among the new PET/CT tracers, radiolabeled choline plays a major role [14, 43], but it can only be recommended at a PSA level over  1 ng/ml [34, 38].

Results of salvage radiotherapy

The PSA level at the time of SRT is the most important predictor for response. In a multicenter study (1,540 patients), the 6-year bNED was 48% with a pre-RT PSA vs. 18% with a PSA level under  0.5 ng/ml, while the overall progression-free survival rate was 32% [53]. A Gleason score of 8–10, pre-SRT PSA over  2 ng/ml, negative surgical margins, postoperative PSA doubling time of less than  10 months, and seminal vesicle invasion were negative prognostic factors. Patients without these features had a 6-year progression-free survival of 69%. Despite a Gleason score of 8–10, patients would benefit from SRT if the pretreatment PSA was less than  2.0 ng/ml, surgical margins were positive, and PSA doubling time was more than  10 months. In this situation, the 6-year bNED was 33% [53]. An earlier treatment start, e.g., with a PSA lower than 0.3 ng/ml, can significantly improve the outcome [48, 49]. Importantly, achieving an undetectable PSA after SRT offers a second chance of cure: In a multivariate analysis of a homogeneously treated group of 162 patients (all pN0, median dose 66 Gy 1.8 Gy fractions), this was the most important predictor for bNED [65]. This is in agreement with earlier data [30].

Total dose of salvage radiotherapy

There is still controversy about the optimal irradiation dose for post-RP SRT/ART. In the guidelines, total doses of “at least 66 Gy” are recommended [23]. However, some recently published series demonstrated a better outcome with higher total doses [3, 26, 40, 49]. In all, 364 prostate cancer patients receiving SRT after RP were followed up for a median 6.0 years. Three dose groups were defined (low: < 64.8 Gy; moderate: 64.8–66.6 Gy; high: > 66.6 Gy). The high-dose group had an improved bNED compared with the low-dose group (HR 0.60) [3]. Similarly, in a retrospective series with 301 patients receiving 66.6 or 70.2 Gy, dose was a significant predictor (p= 0.017 in multivariate analysis) of bNED [49]. Dose escalation seems necessary in patients with histologically confirmed local recurrence.

Salvage radiotherapy plus hormone therapy

In a prospective pilot study, 75 patients were treated with salvage radiation therapy + 2-year ADT. All patients achieved an initially complete PSA response (< 0.2 ng/ml). The relapse-free survival rate at 7 years was 78% [12]. Another study [50] comprised 630 patients (including 66% with high-risk factors) receiving SRT. Twenty-four percent of the patients had concurrent ADT. In high-risk patients, ADT was an independent predictor of progression-free survival.

RTOG 96–01, a randomized phase III trial compared SRT plus ADT with SRT plus placebo in 770 men with pT3/pT2 R1 N0 M0 prostate cancer and an elevated post-RP PSA [47]. Twenty-four months of peripheral androgen blockade during and after RT significantly improved bNED (57% versus 40%; p  <  0.0001) and reduced the incidence of metastatic disease 7.4% versus 12.6%, p  < 0.04) without adding significantly to radiation toxicity. Hence, high-risk patients possibly profit from additional ADT. A current RTOG trial (0534) is investigating the benefit of short-term ADT as well as pelvic nodal irradiation in the SRT setting.

Side effects

The three randomized clinical trials discussed above prospectively collected data on gastrointestinal or genitourinary toxicity. In the EORTC and SWOG trials, 2D treatment planning was used. In the SWOG 8,794 study, 3.3% of the irradiated patients had grade 3 or higher adverse events compared to 0% in the observation group (p = 0.002). The incidence of urethral strictures was higher in the RT group (17.8% vs. 9.5%; p  =  0.02) [57].

In the EORTC trial, there was no significant difference in grade 3 or higher toxicity between the ART and observation groups. In the ART cohort, late grade 2 and 3 toxicity prevailed (p = 0.0005) [5].

In the German study, after 3D-planned RT the incidence of late grade 3 or higher events was only 0.3% [64]. One patient developed a urethral stricture in the observation arm, compared with 2 patients in the ART arm. Urinary incontinence was not assessed in this trial.

In the EORTC study, there was no difference between treatment arms in the number of fully continent men after 24 months [5].

SRT with a dose of 66 Gy is generally associated with a low rate of severe acute and late side effects. Urinary incontinence in 0–5% of the cases, moderate proctitis in 0–10%, and mild to moderate cystitis in up to 10% may result from this procedure [19, 30, 54, 65]. Comparable rates of severe complications were found in an SEER database analysis of 11,522 patients after sole RP [2]. A low rate of side effects is of particular importance for a therapy without histologic confirmation, but it may be difficult to differentiate side effects of RT from preexisting disabilities and sequelae of RP.

Adjuvant radiotherapy vs. salvage radiotherapy

Multiple prospective and retrospective studies analyzed whether ART or SRT is preferable in terms of local control and bNED [3, 5, 26, 27, 30, 49, 53, 57, 60, 64, 65]. A consistently better local control and bNED were observed in ART compared with SRT patients. The 5-year bNED rates are approximately 69–89% after adjuvant radiation therapy. Local control is 96–100% after adjuvant radiation therapy and 79–93% after salvage radiation therapy [7]. In a multicenter study including 449 prostate cancer patients with pT3–4 N0, the 5-year bNED was 73% after ART compared with 50% after SRT (p = 0.007). A Gleason score of 8 was a significant predictor of bNED [59]. These results were confirmed by others [9]; however, the inverse, i.e., an advantage of SRT over ART, has also been reported [31]. For all these reasons, the best choice of treatment (ART vs. SRT) has to be discussed with each patient individually, also considering the risk of overtreatment by immediate postoperative irradiation. Treatment options for patients with undetectable PSA after RP according to the EAU guidelines are summarized in Tab. 3. Our recommendations on the procedures for patients with rising PSA levels after RP are presented in Fig. 1.

Fig. 1
figure 1

Recommended procedures for patients with rising PSA after radical prostatectomy (University Hospital Ulm, Germany). TRUS transrectal ultrasonography, PET positron emission tomography, CT computed tomography

Full size image

In 2007, a prospective randomized study was initiated to address this question as well as the potential role of concomitant androgen deprivation [32]. The RADICALS (Radiotherapy and Androgen Deprivation in Combination After Local Surgery) trial is an effort to evaluate ART versus SRT. Patients are randomized after surgery to early or delayed radiation. Delayed radiation will be given when there are either three consecutive PSA rises or two consecutive rises and a final PSA over  0.1 ng/ml. The planned accrual is 2,600 patients, with cause-specific survival being the primary outcome. There is a second randomization regarding androgen deprivation therapy.

Conclusion

After RP, ART with 60–64 Gy improves bNED rates and potentially also overall survival in high-risk patients. Lacking direct comparative studies, the equivalence of SRT to ART remains to be proven. After PSA relapse, SRT with at least 66 Gy should start as early as possible (with PSA < 0.5 ng/ml). Whole pelvic radiation therapy can be offered as an attractive option for high-risk patients. Modern techniques such as intensity-modulated RT (IMRT) and image-guided RT (IGRT) should be used. Serious side effects are low, confirming the suitability of this approach. According to the RTOG-96–01 trial, peripheral androgen deprivation therapy can improve bNED and reduce the rate of metastatic disease.