Salvage Versus Adjuvant Radiation Therapy Following Radical Prostatectomy in Localised Prostate Cancer: A War Without a Winner

Abstract

Objective: To review current literature regarding the efficacy of adjuvant radiation therapy (ART) and salvage radiation therapy (SRT) following radical prostatectomy (RP) in patients with undetectable postoperative prostate-specific antigen (PSA) levels and high-risk features of prostate cancer (PCa) recurrence. Methods: Seven randomized controlled trials focused on the use of ART compared with either observation or SRT after RP that had been published in PubMed up to May 2022 were reviewed. Results: The use of ART following RP has been the treatment of choice over the past decade. Three RCTs comparing ART with early SRT show that SRT given as soon as biochemical recurrence (BCR) is detected is not inferior to ART while it offers the opportunity to avoid overtreatment and potential RT-related side effects. A meta-analysis summarizing the results from these trials supports these findings. Conclusions: Early SRT may be suggested as the standard of care for patients with PCa and high-risk features for disease recurrence following RP. Nevertheless, further investigations are needed to identify those patients who will benefit from ART, particularly, in case of lymph node involvement. Moreover, some patients might avoid SRT despite reaching detectable postoperative serum PSA levels.

Introduction

Because of its complexity and heterogeneity, prostate cancer (PCa) management at any disease stage is under constant debate[1,2,3], and the use of adjuvant radiation therapy (ART) following radical prostatectomy (RP) to address risk factors for disease recurrence has been questioned in recent years[4].

While earlier randomized controlled trials (RCTs) showed a biochemical recurrence (BCR)-free survival benefit with ART, a meta-analysis published in 2020 has shown no event-free survival benefit compared with salvage RT [5,6,7,8,9]. Nevertheless, the impact of each of the risk factors and their combination on recurrence remain under debate[10].

In this respect, the last European Association of Urology (EAU) guidelines, published in 2022, propose ART be offered to a select group of patients: (a) patients without lymph node involvement but adverse pathology such as International Society of Urologic Pathologists (ISUP) grade group 4–5 and pT3 with or without positive margins and, (b) patients with lymph node metastasized PCa with > 2 nodes involved[11].

In this review, we discuss the RCTs evaluating ART and salvage radiation therapy (SRT) for patients treated with RP who had high-risk features for disease recurrence.

Methods

We searched RCTs and meta-analyses published up to May 2022, using the following terms: “adjuvant radiotherapy, prostatectomy,” salvage radiotherapy, prostatectomy,” and “postoperative radiotherapy, prostatic neoplasm.” Articles not written in English were excluded. Conference abstracts that did not provide enough information, case reports, review articles, and editorial comments were also excluded from review. Seven RCTs focused on the use of ART compared with either observation or SRT after RP were included in our review.

Results

Available evidence

Approximately 25% of men will experience recurrence of their cancer following RP for localised PCa[10]. ART has been the treatment of choice since the publication of 3 RCTs over the past decade: the SWOG 8794[6], the European Organisation for Research and Treatment of Cancer (EORTC 22911)[7], and the German Cancer Society (ARO 96–02/AUO AP 09/95)[8] trials, with their last results published in 2009, 2012, and 2014, respectively.

The SWOG 8794 trial[6] (Table 1) was the first randomized study to support the use of ART in patients with unfavourable pathologic features (pT3 disease or positive surgical margins). Thomson et al. proved that the use of ART leads to better oncologic outcomes in terms of metastasis-free survival (MFS) and overall survival (OS). Nevertheless, it is important to note that the patients enrolled were randomized to ART or observation instead of SRT. Hence, not every patient with BCR received timely SRT. Accordingly, the median serum PSA level at the time of SRT in the case of PSA-only relapse was 0.75 ng/dL, and 37% of patients receiving SRT had already experienced objective cancer recurrence. In addition, 33% of patients included in this trial did not have undetectable postoperative PSA levels.

Similarly, Bolla et al.[7] (Table 1) randomized 1005 patients with pT3 disease or positive surgical margins following RP to ART or a wait-and-see/observation policy. According to the SWOG 8794 trial, ART improved biochemical progression-free survival (PFS) and local control compared with observation. Unfortunately, half of the patients in the wait-and-see arm experienced only BCR (defined as an increase in PSA concentration > 0.2 ng/dL after the lowest postoperative value was measured), but only 14.5% of them received SRT. Thus, at least 30% of patients with BCR after surgery did not receive SRT as the first active salvage therapy. In addition, although pre-SRT PSA values were not reported, the median PSA value before any type of active salvage treatment was 1.7 ng/dL, and 14 patients received SRT despite a PSA level that had not reached a nadir below 0.2 ng/mL after RP.

The ARO 96–02/AUO AP 09/95[8] study (Table 1) included only patients with undetectable postoperative serum PSA levels. The results of this study also favoured ART. ART reduced the risk of progression (biochemical progression, local or distant clinical recurrence, or death from any cause) by 49%. In any case, no patient enrolled in the wait-and-see arm received additional treatment, so we cannot deduce whether the use of SRT could have decreased the differences between study arms.

Following this, many retrospective studies were reported that focused on the effect of SRT given to those patients with unfavourable clinical and pathologic features[10]. Aiming to shed light on this topic, Tao et al.[10] published a systematic review and meta-analysis that included 15 studies published between 2002 and 2018 comparing the prognosis of ART and SRT. It must be noted that this review did not include the above 3 RCTs as those RCTs compared ART with observation, while the review focused on ART versus SRT. Therefore, all studies considered for analysis were retrospective. According to this meta-analysis, the use of ART following RP reduces the risk of BCR and 5-year overall mortality, although no benefit from ART was demonstrated in terms of patient survival at 10 years from RP. Importantly, (a) all patients considered had received either ART or SRT, which means that all patients included in the SRT group had developed BCR, which in turn may imply a greater predisposition to unfavourable disease, and (b) no improvement in BCR in the ART arm was observed when only early SRT (defined as SRT given at PSA levels < 0.5 ng/dL) was considered.

Finally, a Finnish trial published in July 2019 (Table 1) randomized 250 patients with pT2 disease and positive margins or pT3a disease (irrespective of margin status) to ART or observation[9]. ART prolonged BCR-free survival compared with RP alone, but (a) the increase in PSA levels before receiving SRT was counted as an event, while nearly three-quarters of patients achieved PSA remission after SRT; (b) 29% of patients in the observation group whose disease progressed locally did not receive SRT; and (c) the median PSA level at SRT was 0.7 ng/mL (range, 0.42 to 8.2).

In short, the above-mentioned studies support the use of ART when it is compared with observation or salvage radiotherapy, but they do not provide adequate outcome information about early SRT.

In this vein, several retrospective studies have suggested that early initiation of SRT will not compromise cancer control but rather reduce the overtreatment that is associated with ART[12,13,14,15]. In 2020, a meta-analysis showed that SRT results may not differ from ART results if the postoperative serum PSA value is below 0.2 ng/mL at the time of treatment[5]. Consequently, it has recently been disputed whether the use of SRT as soon as PSA values reach a detectable level may have an important impact when SRT is compared with ART.

New investigations and findings

At present, SRT is recommended when the postoperative serum PSA value is ≤ 0.5 ng/mL[13,16]. It has been observed, however, that 5-year freedom from BCR after SRT was 71% for those patients with a pre-SRT PSA level ≤ 0.2 ng/mL, 63% for those with a PSA of 0.21 to 0.50 ng/mL, 54% for those with a PSA value between 0.51 and 1.0 ng/mL, and 37% for those with a PSA value ≥ 2.0 ng/mL[17]. Hence, the timing of SRT may alter its effectiveness[18].

Three RCTs — RADICALS-RT, RAVES, and GETUG-AFU-17 — were aimed at answering a hitherto unresolved dilemma: “What is the best timing for postoperative RT?”[19,20,21] (Table 1). In this case, patients were randomized to ART or observation followed by SRT when the PSA level was ≥ 0.2 ng/dL[19,20] or even ≥ 0.1 ng/dL[21]. A collaborative and prospectively designed systematic review and meta-analysis that included these 3 trials suggested that ART does not improve event-free survival in men with localised or locally advanced PCa compared with SRT. Five-year event-free survival in ART and SRT were 89% and 88%, respectively[5]. Of interest, more than half of the patients included in the SRT group did not meet the BCR criteria avoiding overtreatment and potential RT-related-side effects, including sexual, urinary, and bowel dysfunctions[5,22]. The RADICALS-RT, RAVES, and GETUG-AFU-17 trials[19,20,21] revealed that RT toxicity was more common among patients randomized to ART compared with SRT. Grade 1 to 2 events were about twice as prevalent in the ART group; grade 3 to 4 events were uncommon in both groups.

Data from these RCTs support the notion that early SRT avoids overtreatment without compromising the oncologic outcome. Therefore, it seems reasonable to adopt this approach as the standard of care following RP for localised prostate cancer. Nevertheless, can we generalise it to all patients? Should we give special consideration to the different risk factors for recurrence and their combination? Will the development of biomarkers and imaging techniques allow us to identify the best timing for RT after surgery?

Table 1. Adjuvant radiation therapy randomized controlled trials.

Table 1. Adjuvant radiation therapy randomized controlled trials.

Discussion

Accurate risk of recurrence estimation

Extracapsular extension, seminal vesicle invasion, positive surgical margins, high Gleason score, and lymph node involvement are the main independent predictors of BCR[10,23,24]. However, the increasing use of genomic biomarkers to identify both germline and somatic variations may provide a more precise assessment of patients at risk of recurrence[25,26].

The heterogeneity of PCa, with variable responses to treatment, is probably related to the molecular heterogeneity of this disease, which in turn is closely related to its genetic profile. Thus, germline testing together with molecular profiling or tumour genomic profiling may lead us to understand more about the biology of the tumour, and distinguish PCa with indolent behaviour from those cases with a lethal course[27]. Approximately 5% of patients with localised PCa harbour germline variations[28,29], such as BRCA2, and ATM variations that are associated with both PCa susceptibility and higher risk of aggressive disease[28]. In this regard, BRCA1/2 carriers with localised PCa have a 16% increased absolute risk of developing metastases[30]. Therefore, the identification of these variations may have a role in the management of patients at risk of BCR after RP.

Somatic changes in DNA repair genes are found in nearly 10% of PCa tumours confined to the prostate gland[31]. Spratt et al. evaluated a 22-marker genomic classifier (GC; DatabasE of genomiC varIation and Phenotype in Humans using Ensembl Resources [DECIPHER], Sanger Institute) to predict metastasis through a metanalysis of 5 studies. The authors proved that GC was an independent predictor of metastases. Likewise, the 5-year cumulative incidence of metastasis was 2.4% for patients with a low GC score but 15.2% for patients with high-risk GC scores[32]. Later, Marascio et al. explored the clinical benefit of decision-making based on DECIPHER CG testing after RP. This prospective observational study revealed that patients with high GC risk who received ART had a 2-year cumulative incidence of PSA recurrence 8 times lower than those who did not.

How should patients with very high-risk features and a combination of high-risk features and undetectable postoperative serum PSA levels be treated?

Subgroup analysis in the ARTISTIC meta-analysis was limited by the low event rate, while the effect of RT timing could not be properly evaluated in patients with high-risk features for BCR such as seminal vesicle involvement (20% of patient enrolled), Gleason score ≥ 8 (15% of patients enrolled) or node-positive disease (3% of patients enrolled)[5]. Hence, there are still doubts about how to manage those patients with undetectable postoperative PSA levels together with high-risk features, such as node-positive disease, Gleason score > 8, and/or pT3b disease or higher[4]. In addition, investigators in the EORTC 22911 trial determined that patients with 2 risk factors (eg, pT3a-b plus positive surgical margins) tended to have a greater risk of BCR and death if they received ART compared with those who presented with a single risk factor[7]. No combination of adverse pathologic features was evaluated in the ARTISTIC meta-analysis[5], preventing us from knowing which is the best management option for these patients.

Particularly controversial has been the management of pathologically node-positive PCa patients given the prognostic variability according to the number of affected nodes and tumour characteristics. In this regard, Abdollah et al. suggested that patients with high-risk features for BCR (pathological stage > T3a and/or positive surgical margins together with Gleason score 7 to 10) may benefit from ART even when the number of lymph nodes involved is lower than 3[33,34].

In a systematic review of 26 studies published in 2018, only one propensity score-matched study provided information about the effectiveness of initial observation and SRT versus ART in the pN1 PCa setting[34,35]. The study proved a benefit in terms of 4-year MFS for those patients receiving ART regardless of disease characteristic[35].

Recently, Tilki et al., showed that ART may be better when compared to SRT. Moreover, the benefit seems to be directly related to the number of lymph nodes affected. Therefore, ART might be avoided when ≤ 3 positive lymph nodes are found (notably if at least 12 lymph nodes are sampled) but the impact that combination of positive lymph nodes with other risk factors for BCR may have on oncological outcomes was not explored[36].

In summary, RCTs will be needed to determine the benefit of ART in positive lymph node PCa patients. Until then, the use of ART should be always considered in this group of patients, especially if more than 2 lymph nodes are involved and/or other risk factors for BCR are associated.

Impact of novel imaging techniques on treatment

The detection rate of different imaging modalities is determined by PSA level. Thus, the lower the PSA value, the lower the possibility of identifying locoregional or distant disease[37,38]. The decision to initiate ART after RP should be made when serum PSA values are < 0.5 ng /mL (or ideally < 0.2 ng/mL); therefore, the use of sensitive and accurate imaging techniques is of special interest. Choline-based positron emission tomography (PET)/computed tomography (CT) detection rate is 30% when the serum PSA level is ≤ 1 ng/mL after RP, while the corresponding detection rate is 60% for gallium 68 prostate-specific membrane antigen (68Ga-PSMA) PET/CT. Furthermore, when serum PSA is below 0.2 ng/mL the estimated detection rate for 68Ga-PSMA) PET/CT is 40%[38,39]. PET imaging with 68Ga-PSMA ligand was approved on December 1, 2020, by the US Food and Drug Administration to detect PCa recurrence after RP[40].

A meta-analysis of 15 studies showed that 68Ga-PSMA PET results after RP influenced the type and characteristics of treatment. The number of patients who received an increased dose or target volume of SRT rose after 68Ga-PSMA PET[41]. Similar results were presented at ASCO 2022. Armstrong et al. randomized 193 patients who experienced BCR to undergo any conventional imaging or 68Ga-PSMA-11 PET/CT scan prior to SRT (median PSA before SRT of 0.3 ng/mL). Seventy-one percent of major changes, defined as change of androgen-deprivation therapy (ADT) duration ≥ 3 months, change of standard RT volumes, target volume delineation beyond standard RT field, simultaneous-integrated boost beyond standard RT fields, and initiation of advanced systemic therapy, were PSMA-related[42].

Pending final results of the Radiation Therapy Oncology Group (RTOG) 0534 SPPORT trial, which explores the effect of ADT with or without pelvic lymph node treatment added to the prostate bed together with SRT[43], novel imaging techniques may be useful in identifying patients who may benefit from whole-pelvis SRT, although its association with improved oncologic outcomes is not yet known[44].

RT technique

RT techniques in PCa have advanced significantly in the past decades. For example, intensity-modulated RT (IMRT) is now the gold standard for external beam RT (EBRT)[5]. Although the benefit of dose escalation to 78 Gy or higher for definitive localised PCa RT is well known, its effect in the postoperative setting is still in doubt. A systematic review and meta-analysis of 71 retrospective studies demonstrated a proportional gain in BCR-free survival of 2% per incremental Gy when escalating SRT > 70 Gy was given. A randomized phase 3 study that randomized patients with BCR following RP to either 64-Gy or 70-Gy SRT showed that dose escalation had a minor impact on quality of life but led to significant worsening of urinary symptoms[45]. Thus, neither National Comprehensive Cancer Network (NCCN) nor EAU guidelines specify the appropriate dose to be administered and recommend doses of at least 66 Gy up to 72 Gy[11,46].

Stereotactic body RT (SBRT) that accurately delivers a high radiation dose to an extracranial target in 1 or a few treatment fractions (extreme hypofractionation)[47] has been suggested as a safe procedure for patients with BCR and detectable local recurrence. Although further investigations are needed, this emerging treatment option for isolated relapse may reduce the undesirable side effects of RT without compromising oncologic outcomes[48].

None of the studies included in the ARTISTIC meta-analysis administered RT doses higher than 66 Gy or used hypofractionation techniques. Therefore, new advances in RT may still improve oncologic outcomes following SRT.

Timing and duration of ADT with SRT

The optimal duration of ADT together with SRT is uncertain. The RTOG 9601 trial supports the addition of 2 years of daily bicalutamide 150 mg to SRT because a benefit in OS at 10-year follow-up was observed (16% reduction in risk of death)[49]. Similarly, the results of the GETUG-AFU 16 trial, which randomized 743 patients to SRT alone or SRT plus 6 months of quarterly goserelin, confirmed that the addition of ADT led to a 27% reduced risk of metastasis[50]. Fossati et al. retrospectively explored the impact of ADT duration on oncologic outcomes following SRT and concluded that patients with more than one risk factor (≥ pT3b, ISUP > 3, or PSA level at SRT > 0.5 ng/mL) may benefit from long-term ADT, whereas patients with a single risk factor may receive < 12 months of ADT without compromising oncological outcomes. Patients without any risk factors did not show a significant benefit from concomitant ADT[51]. Irrespective of the study arm, some patients enrolled in the RADICALS trial were randomized to 0 versus 6 versus 12 months of ADT, but results regarding ADT effect are still unknown[21]. Likewise, the ongoing LOBSTER trial investigates whether prolonging the duration of ADT from 6 to 24 months improves oncologic outcomes. Result from these 2 RCTs will allow us to define the most appropriate treatment time for patients receiving SRT.

Several ongoing RCTs are evaluating the role of new androgen receptor pathway targeting agents in combination with SRT. For instance, the SALV-ENZA trial[52] compares SRT plus placebo with SRT plus enzalutamide, while the STEEL trial[53] compares SRT plus standard ADT with SRT plus standard ADT plus enzalutamide. The CARLHA-2 trial is exploring the effect of adding apalutamide to SRT and standard ADT[54]. Finally, the FORMULA-509 trial is studying the addition of apalutamide, abiraterone, and prednisone to SRT plus standard ADT[55].

Benefit of SRT in cases of BCR

Lastly, which patients would truly benefit from SRT in the case of BCR? Beyond reporting the role of ART in the presence of high-risk features for BCR, the SWOG 8794[6],EORTC 22911[7], ARO 96–02/AUO AP 09/95[8], and Finnish[9] trials provided information about the natural history of PCa. They revealed that median MFS and clinical PFS rates were > 10 years even among men included in the observation arm. These findings raise the question of whether all patients with BCR would benefit from SRT and suggest that some patients with BCR following RP who receive SRT may be overtreated. For instance, Pak et al. retrospectively analysed 817 patients with BCR after RP. Patients were categorised into 3 groups according to time from RP to BCR: an early group (median BCR-free survival of 8.5 months), an intermediate group (median BCR-free survival of 17.5 months), and a late group (median BCR-free survival of 70 months). The authors found that 8-year distant MFS and 8-year CSS were significantly better among patients receiving early SRT (with or without ADT) than among those receiving ADT alone in cases of early BCR, but no differences were found in cases of late BCR. Because patients included in the early BCR group had worse clinical and oncologic characteristics, such as higher NCCN risk group, preoperative PSA > 20 ng/mL, Gleason score 9 to 10, and pT3b-stage disease, more challenging tumour biology may explain these results. Nevertheless, not all studies published to date support time from RP to BCR as an independent risk factor for worse oncologic outcomes[56,57].

Although PSA doubling time (PSADT) is not always associated with short BCR-free survival, it has been noted that these variables are interconnected[56]. Thirty-six retrospective studies evaluating prognostic factors in patients with BCR following RP that were included in a systematic review and subsequent meta-analysis showed a significant relationship between shorter PSADT and higher risk of distant metastasis, PCa-specific mortality, and overall mortality, whereas BCR-free survival was also related to PCa-specific mortality and overall mortality.

In contrast, after PSADT, PCa Gleason score based on prostatectomy histology report was recognized as a second factor more strongly associated with worse oncologic outcomes following RP as a primary treatment. Based on these findings and the hypothesis that not all patients with BCR benefit from salvage treatment, the EAU proposed a new BCR risk stratification that suggested that EAU low-risk BCR patients after RP (PSADT > 1 year and pathologic Gleason score < 8) might avoid salvage treatment[57]. Nevertheless, a recent validation showed that SRT was highly protective, with maximum effect in early delivery[58]. Therefore, risk group stratification after BCR remains inaccurate, and more research is needed to develop a molecular approach.

So far, only retrospective studies have been published regarding this topic. When the ART versus SRT debate comes to an end, selecting those patients who may benefit from SRT seems to be the next step. In this regard, RCTs are of the utmost importance for shedding light on the hypotheses retrospective studies have raised[59,60].

Conclusions

Early SRT may be suggested as the standard of care for patients with PCa with high-risk features and undetectable postoperative PSA. Nevertheless, some aspects, such as the duration of ADT with SRT as well as a more accurate stratification of patients at risk of clinical progression, require further investigation. Therefore, some patients with undetectable PSA following RP may benefit from ART, while other patients can avoid SRT despite reaching serum PSA levels above 0.2 ng/mL. In conclusion, only improving patient selection for ART or SRT will lead to peace.