7. Diagnosis and treatment of recurrent or metastatic prostate carcinoma – Prostate Cancer

Background information

Re: Recommendation 7.1

The diagnosis of a recurrence is primarily based on the detection of a PSA increase, i.e. biochemically. A PSA rise following radical prostatectomy or radiotherapy with curative intent may be due to either local recurrence or metastases. The localization of the recurrence is only of crucial importance if treatment for the local recurrence is an option for the patient. If this is ruled out because of a high risk of complications, serious concomitant diseases, or the patient's refusal to undergo this measure, no diagnosis is necessary because it is of no consequence. Over the further course of the disease, examinations should be symptom-oriented.

Re: Recommendation 7.2

Several studies have addressed the question of the PSA level at which recurrence is clearly present following radical prostatectomy. There have been investigations into the PSA values at which the following are probable: further increases in PSA over time; local recurrence; or metastasis. Since PSA values in the range of the detection limit do not necessarily have to increase further (e.g. in the case of residual benign glands) and measurement inaccuracies can occur in the low-PSA range, an increasing PSA value of > 0.2 ng/ml, measured at an interval of at least two weeks, confirmed in at least two measurements, was determined to be a suitable definition for significant PSA recurrence. This recommendation is identical to the current version of the European Society of Urology guidelines. However, two rising values < 0.2 ng/ml can also be taken to mean biochemical recurrence [740], [741], [742].

Re: Recommendation 7.3

The previous ASTRO definition of PSA recurrence after radiotherapy - three consecutive increases - was widely criticized: it was not of sufficient predictive value with regard to further clinical progression or overall survival, nor was it of reliable value in patients currently undergoing/having undergone hormone-ablation therapy. Moreover, there was a possibility of bias in disease-free survival rates calculated according to Kaplan-Meier by backdating the timing of PSA progression [743]. Among the alternative models investigated and retrospectively evaluated in large patient groups, the ASTRO definition is now: baseline plus 2 ng/ml ("Phoenix definition"). This was shown to be more sensitive and specific in detecting PSA recurrence after primary radiotherapy [748]. The dating of the recurrence is thereby determined "at call", i.e. at the documented time of exceeding the threshold (the second measurement is to be taken at an interval of approximately three months). This definition is applicable following teletherapy and brachytherapy, and has also proven to be valid in case of additional (neo)adjuvant short-term hormone-ablation therapy and/or shorter follow-up times [749], [750]. Compared to the previous definition, the "Phoenix definition" is significantly less sensitive with regard to false positive findings, which can be caused, for example, by the so-called "PSA bounce" (= reversible PSA increase after radiotherapy) [751].

Re: Statement 7.4

Several studies have shown that biopsy of the vesico-urethral anastomosis is an inaccurate method for detecting local recurrence [744], [745]. Since this is especially true for the low PSA range (therapy of PSA recurrence).

Re: Recommendation 7.5

Prior to local therapy, the diagnosis of a local recurrence should be confirmed by transrectal or transperineal punch biopsy. The histological specimen should be evaluated by a pathologist familiar with pathomorphological changes following radiotherapy in order to avoid false positive findings [747]. The biopsy should be performed once any PSA bounce phenomenon (= reversible PSA increase after radiotherapy) has been ruled out [752], [753], i.e. only in the case of a confirmed PSA recurrence (see chapter Definition and diagnosis of tumor recurrence).

Re: Recommendation 7.6

The most important factors for differentiating a local recurrence from a systemic recurrence are the PSA doubling time, the Gleason grade of the prostatectomy specimen and the time interval between surgery and recurrence detection. Systemic progression is indicated by a short PSA doubling time (e.g., B.)  7 makes systemic progression likely in recurrence. The shorter the time interval between PSA recurrence and detectable PSA level, the more likely systemic progression is present [661], [754].

Re: Recommendation 7.7

The probability of metastasis in relation to the PSA level in cases of recurrence following primary local treatment with curative intent is not known with any precision. In cases presenting following primary radiotherapy and/or brachytherapy with proven recurrence, metastasis should therefore be ruled out as being unlikely by means of the existing available options before any possible salvage surgery. In the absence of clear data on this issue, a PSA threshold of 10 ng/ml is considered to warrant bone scintigraphy. 

In cases having undergone radical prostatectomy where salvage radiotherapy is intended, this will be indicated anyway at low PSA values. The available data show that bony metastasis virtually does not occur at PSA < 7 ng/ml, is very rare at PSA < 10 ng/ml, and only at a PSA level of 20 ng/ml or higher do nuclear medical examinations show positive findings with sufficient probability. Therefore, the recommendation of a threshold of 10 ng/ml is a realistic compromise in order to avoid unnecessary examinations [755]

The following recommendations and statements refer to recurrence assessed as localized. For any suspected distant metastasis, see chapter 7.3 onwards.

In addition, by way of introduction, please refer to chapter 5.3 Staging, in particular Recommendation 5.29 on diagnosis of recurrence regarding the use of PSMA-PET-CT in the diagnosis of PSA progression following radical prostatectomy.

Summary of the evidence base on radiotherapy for PSA recurrence following radical prostatectomy.

A recently published systematic review is available [756]. Randomized trials of radiotherapy for progression following radical prostatectomy (so-called salvage radiotherapy, SRT) which include comparison with a control group do not exist. For SRT, a controlled study is available [757], in which SRT was retrospectively compared with a "wait-and-see" strategy. It showed a carcinoma-specific survival advantage for SRT, but no survival advantage [757]. However, methodological flaws diminish the value of this study. In addition, two randomized trials are available in which SRT alone was compared with SRT+short-term hormone-deprivation therapy (GETUG-AFU 16) [758] and SRT compared with SRT+long-term therapy with bicalutamide 150 mg / d (RTOG 96-02) [759]. In these two randomized trials, the RTOG trial showed a significant survival benefit (primary endpoint) for combination therapy, and the GETUG trial showed an improvement in metastasis-free survival over 6 months with short-term ADT [759][758]. Thus, the value of SRT as a curative therapy can be considered to be well established.

The natural history of PSA progression following radical prostatectomy (RP) can be long-term even without any treatment intervention, whereby patients do not develop clinically significant symptoms. In a large retrospective series, approximately 2,000 patients were followed up for a median of 5.3 years (0.5-15 years) after RP [661]. 15 % of patients (n=315) developed PSA recurrence [661]. Of these, distant metastases occurred in 103 (34%) during the study period [661]. The median time between PSA recurrence and the diagnosis of distant metastases was 8 years. Among patients with distant metastases, these led to death after a median of 5 years [661]. Based on the data collected, the expected rate of patients with metastasis-free progression after 15 years was calculated to be 82% [661]. Overall, however, this is a favorable patient population with a relatively low proportion of patients at high risk of progression [661]. Only 6% had lymph node involvement, 5 % had seminal vesicle infiltration, and 93% of patients had a Gleason score of max. 7 [661]. It can therefore be assumed that significantly higher rates of PSA progression occur in collectives with more advanced tumors.

Salvage radiotherapy is a therapeutic option with curative potential. However, as an invasive procedure, it is also associated with potential adverse side effects. In this regard, it should be kept in mind that the rate of severe late effects > RTOG 2 (Radiation Morbidity Scoring Criteria) is comparatively low with modern RT technology (Intensity-Modulated Radiation Therapy – IMRT) [674][758]. In the series described [661], patients with favorable prognostic criteria (PSA doubling time > 10 months, occurrence of PSA recurrence later than 2 years after RP, primary Gleason score < 8) had significantly less PSA progression than those with unfavorable prognostic criteria.

In the systematic review on salvage radiotherapy cited [756], patients with a PSA doubling time of < 12 months and a Gleason score > 8 had significantly worse overall survival with biochemical progression without therapy than those with a doubling time of > 12 months and a Gleason score < 8 [756]. Accordingly, two risk groups (European Association of Urology biochemical recurrence risk groups) were defined. This classification was externally validated by a retrospective study of 1,125 patients from the Martini-Klinik in Hamburg, Germany, treated between 1992 and 2006 [760]. This showed that both cancer-specific survival and metastasis-free survival were independent predictors of subsequent outcome [760]. This systematic review and external validation form the basis of the EAU Prostate Cancer Guidelines Panel recommendations for salvage radiotherapy [761].

This work forms the background of the formulated statement. It is important to note here that many additional points must be included in this decision such as patient age, the level of PSA prior to possible salvage radiotherapy, and the patient's concomitant diseases and life expectancy. It must always be kept in mind that from a PSA level of > 0.5 ng/ml prior to initiation of SRT, the curative potential is significantly reduced or lost.

Treatment of metastatic (M1), hormone-sensitive prostate carcinoma has fundamentally changed in recent years. After the CHAARTED study showed an advantage for early combined chemohormonal therapy [788], this could later also be shown for combination therapies of androgen deprivation (AD) with the new generation of hormone drugs (abiraterone, enzalutamide, apalutamide) [790][786][798][792][791]. Corresponding to what is seen with simple androgen deprivation, individual studies that also included locally-advanced high-risk tumors (e.g., STAMPEDE) suggest that patients at earlier tumor stages may also see a similar benefit from the combination therapies.

Re: Recommendation 7.20

In the trials that demonstrated a benefit from combined hormone therapy with docetaxel or abiraterone (plus prednisone/prednisolone) in hormone-sensitive metastatic prostate cancer, the study populations were defined in different ways based on metastatic pattern or Gleason score. For combined chemohormonal therapy (AD+docetaxel), patients were divided into high and low volume (CHAARTED) [788]. High volume includes those patients in whom the following were present:

·         at least four bone metastases, including at least one outside the axial skeleton or pelvis

·         and/or visceral metastases.

Only patients who had de novo metastatic disease and who met the following high risk criteria were included in the pivotal study with abiraterone (LATITUDE) [786][798]:

At least two of the three parameters:

·         Gleason 8 – 10

·         at least three bone metastases

·         visceral metastases.

Studies in which patients were classified according to both high and low volume as well as high and low risk show a little more than 80% agreement between high volume and high risk. Nevertheless, differentiation or classification of patients with metastatic hormone-sensitive prostate cancer into both groups is reasonable, as clear advantages for chemohormonal therapy have especially been shown in high-volume patients, and the approval of abiraterone is limited to high-risk patients.

Re: Recommendation 7.21

The recommendation on patient education is set against the background of informed decision-making as it corresponds to good clinical practice as well as the requirements of the Patient Rights Act. Patients should make decisions together with the informing physician regarding the difficult question of weighing up the risks. The typical and frequent side effects of androgen deprivation and, if applicable, combined chemotherapy or new hormone therapy listed in Table 18 and Table 19 (see Chapter 7.7) are part of the informed consent discussion.

Re: Recommendation 7.22 a-e and 7.24 a-e

The following paragraph summarizes the state of the literature for the combinations of AD and docetaxel or the new generation hormone drugs (abiraterone, apalutamide and enzalutamide) in mHSPC. For the combination hormonal therapies, the recommendations are based on the respective pivotal trials and the resulting regulatory texts; these take into account the patient populations included, some of which differed considerably in the trials and are therefore presented in detail in each case below.

For example, although the data from the STAMPEDE trial suggest that abiraterone is effective beyond the patient population with high-risk criteria, the guideline experts did not see a need to make a recommendation beyond the approval in view of the numerous treatment alternatives. Accordingly, Recommendations 7.19 and 7.21 take into account the indication and range of treatment approaches of the pivotal studies, for which data from RCTs are accordingly available.

Recommendations 7.24a (apalutamide) and 7.24b (enzalutamide) each consider a period of 3 months between the onset of AD and the initiation of apalutamide or enzalutamide therapy. However, in the respective studies, a period of 6 months was allowed for androgen therapy, provided that patients were still receiving docetaxel chemotherapy (up to 6 cycles) before the initiation of the study treatment. The proportion of patients with prior docetaxel chemotherapy was 11 and 18%, respectively, such that the subgroup analyses do not allow us to determine whether additional chemotherapy before starting the new androgen receptor blocker is appropriate.

The duration of therapy is long-term for all hormone combination therapies, but according to good clinical practice, it should be discontinued or modified in case of disease progression or the occurrence of intolerable side effects.

Re: 7.22a and 7.24a (apalutamide)

The combination of apalutamide in addition to AD was studied in TITAN, a randomized, double-blind, placebo-controlled phase 3 trial with a total of 1,052 patients [791]. Patients with documented mHSPC with an ECOG performance status of 0 or 1 who had received prior AD of up to 3 months in duration or up to six cycles of docetaxel with AD of up to 6 months in duration were included [791]. Patients with metastatic disease after local therapy with curative intent (e.g., prostatectomy, local radiation therapy) were also included in the study, provided it had been completed at least one year before randomization [791]. Excluded patients included those with severe angina pectoris, myocardial infarction, heart failure, thromboembolic events, or ventricular arrhythmias [791].

Patients in the intervention group received apalutamide in addition to AD, and patients in the control group received placebo in addition to AD [791]. Primary endpoints were radiographic progression-free survival (rPFS) and overall survival (OS); secondary endpoints were time to chemotherapy, time to worsening pain, time to long-term opioid use, and time to skeletal-associated events [791].

Overall, there is a significant benefit for the combination of apalutamide + AD for the primary study endpoints of rPFS and OS. The total number of adverse events (AEs) and grade 3-5 AEs differed only slightly between intervention and control groups. Skin rashes were observed significantly more frequently with apalutamide. A summary of the results, including adverse events, can be found in Table 15 below [790][791].

Apalutamide for the indication of mHSPC was also subjected to a benefit assessment according to §35a SGB V by IQWIG. As part of this procedure, the G-BA defined the appropriate comparative treatment (aCT) as chemohormonal therapy with docetaxel or, for patients with high-risk criteria, abiraterone. In the absence of a direct comparative study between aCT and apalutamide, an adjusted indirect comparison (AD as "bridge comparator") was implemented to compare apalutamide with docetaxel data from the STAMPEDE trial. Overall, no statistically significant difference in overall survival was found, but a positive effect was shown for apalutamide in terms of adverse events. Overall, IQWIG therefore saw evidence of an additional benefit that could not be quantified [799][800]. The G-BA, on the other hand, considered an additional benefit as not proven, since an advantage for apalutamide was only seen for the endpoint of serious adverse events in the indirect comparison and this did not allow an additional benefit to be determined with sufficient certainty [801][802].

Re: 7.22b and 7.24b (enzalutamide)

The combination of enzalutamide in addition to AD versus AD alone was studied in ARCHES, a randomized, double-blind, placebo-controlled phase 3 trial involving a total of 1,150 patients with pathologically confirmed mHSPC [792]. Adult patients with an ECOG performance status of 0 or 1 who received prior AD of up to 3 months or up to six cycles of docetaxel with AD of up to 6 months pretreatment were included [792]. As in the TITAN trial (apalutamide), patients with metastatic disease after local therapy with curative intent (e.g., prostatectomy, local radiation treatment) were eligible for study participation. Patients with clinically significant cardiac disease (e.g., myocardial infarction within 6 months prior to inclusion) were excluded [792]. The primary endpoint assessed was radiographic progression-free survival (rPFS) [792]. Secondary endpoints were time to PSA progression, time to initiation of new antineoplastic therapy, PSA rate below detection limit, objective response rate, time to worsening of symptoms related to the urinary tract, overall survival (OS), time to skeletal-related symptoms, time to castration resistance, time to worsening of quality of life, and time to worsening of pain symptoms [792].

After a median observation time of 14.4 months, there was a significant benefit for the combination of enzalutamide + AD for the primary study endpoint of rPFS. An OS benefit was not demonstrated in this data set (October 14, 2018) (HR 0.81; 95% CI 0.53-1.25) [792]. The total number of adverse events (AEs) and grade 3-5 AEs differed little between intervention and control groups. A summary of the results, including additional adverse events, is provided in Table 16 below.

For time to worsening quality of life, one of the secondary endpoints, a median of 11.3 months is reported for the intervention group and a median of 11.1 months for the control group (HR 0.96 (0.81-1.14) p=0.6548) [792].

Re: 7.22c and 7.24c (abiraterone)

Data on the combination therapy of abiraterone and AD are available from two randomized clinical trials, STAMPEDE and LATITUDE [798][803].

Only patients in good general health (ECOG 0-1) were included in the STAMPEDE trial (whereas LATITUDE also included ECOG 2). Similar to the docetaxel arm of STAMPEDE (see 7.19 d), there was no difference in metastatic burden in the metastatic group in the abiraterone arm. In terms of overall survival (OS), there was a statistically significant advantage for the combination therapy with a hazard ratio (HR) of 0.63 (95% CI 0.52-0.76) [803]. There was also a significant advantage in failure-free survival between the treatment and placebo groups (HR 0.29; p<0.001) [803] (see Table 17).

The pivotal LATITUDE study included patients with newly-diagnosed, pathologically confirmed mHSPC (high risk) and ECOG 0-2 who had at least two of three designated risk factors (Gleason score of 8-10, at least three bone lesions, measurable visceral metastases) [798][786].

Patients who received prior chemotherapy, radiotherapy, or surgery for the treatment of metastatic prostate cancer were excluded, with the exception of AD with luteinizing hormone releasing hormone (LH-RH) analogues for a duration of less than three months, orchidectomy, or palliative radiotherapy or palliative surgery for symptomatic treatment, respectively [786][798].

The primary endpoint considered was OS and radiographic progression-free survival (rPFS) [786][798]. The evaluation was performed as part of an initial interim analysis in October 2016, the results of which resulted in unblinding as well as a protocol adjustment to an open-label extension phase of the study with the possibility of a change in treatment group in February 2017 [786][798]. In the final evaluation, the change of patients from the control group to the intervention group after unblinding (crossover, n=72) has to be considered [786].

The final evaluation was performed as of August 15, 2018, with a median observation time of 51.8 months [786]. The benefit in terms of overall survival was confirmed (HR 0.62 and 0.66, respectively, see Table 17). Evaluation of rPFS was only performed for the first interim analysis showing a clear advantage of combination therapy at 33.0 vs. 14.8 months, HR 0.47 [786][798].

The side effects in Table 17 are taken by example from the STAMPEDE study.

In accordance with the treatment regimen of the studies [798][803], abiraterone is recommended at the dosage of 1000 mg/day plus prednisone or prednisolone 5 mg/day starting within 3 months after initiation of androgen deprivation.

Re: 7.22d, e and 7.24d, e (docetaxel)

Three randomized clinical trials and a meta-analysis [804] were identified to assess the effectiveness of combination therapy of docetaxel and AD. Two of three trials evaluating combination therapy of docetaxel with concurrent AD showed significant prolongation of overall survival (OS) by 15 and 13.6 months (60 vs. 45 and 57.6 vs. 44 months; 2,962 and 790 patients, respectively) [788][794]: differences in outcomes in a third trial (OS 62.1 vs. 48.6 months; 385 patients) were not statistically significant [793]. Progression-free survival (PFS) or survival without treatment failure was significantly prolonged by combination therapy in all three trials (progression: by 10 and 8.5 months, respectively; treatment failure: by 17 months). Two of three studies (CHAARTED and GETUG) performed a subgroup analysis for patients with high tumor burden (defined in both studies as "visceral metastases or ≥4 bone lesions with ≥ 1 beyond vertebral bodies and pelvis", in GETUG only as post-hoc analysis) and find significantly better results for this subgroup. The study with the largest population (STAMPEDE) does not perform this subgroup evaluation a priori and comes to a significant result for the whole group. In a retrospective analysis of imaging at study inclusion with a subsequent classification according to high and low volume criteria (corresponding to CHAARTED), an evaluation with regard to metastatic burden was also performed for the STAMPEDE study. In contrast to the CHAARTED study, there was no heterogeneity between the treatment groups with different tumor volumes (interaction p-value: 0.827), i.e., according to the authors, the benefit of chemotherapy was independent of tumor burden.

None of the three studies performed subgroup analyses regarding symptomatic versus asymptomatic patients. Given the restrictive inclusion criteria of the trials and prognostically favorable factors such as a median age of 63.5-65 years and the increased grade 3-5 toxicities observed in all trials in the respective docetaxel arm, the recommendation is for patients in good general health with ECOG scores of 0 or 1 in whom the metastatic pattern meets high-volume criteria.

Conflicting data are available for low-volume metastasis. While in some studies or meta-analyses no influence is seen for the "metastatic volume" on the efficacy of early chemohormonal therapy, in other evaluations, no advantage for low-volume patients can be demonstrated [805][806]. In view of the numerous treatment alternatives now available and the toxicity profile of chemohormonal therapy compared with combined hormone therapy, the recommendation for low-volume metastasis has been weakened ("can").

The dosing recommendation of 75 mg/m² every three weeks for six cycles for docetaxel administration in chemohormonal therapy is consistent with the dosing and prioritized frequency used in RCTs addressing this question [788][794] and has thus found its way into the European Medicines Agency's SmPC. In contrast to metastatic castration-refractory prostate cancer (cf. Recommendation 7.39), docetaxel administration at 50 mg/m² every two weeks at the metastatic hormone-sensitive stage has not been investigated in prospective studies and is therefore not recommended. According to the longest time period available in the included studies, the start of chemotherapy is recommended no later than 4 months after the start of androgen deprivation.

Re: Recommendation 7.23

There are now numerous combination therapies for mHSPC for which superiority over AD alone has been demonstrated. The various therapies differ, among other things, in terms of approval status or indication as well as spectrum of side effects. No statement can be made regarding the advantages of one or the other variant of combination therapy for specific patient groups, as there are no directly comparative studies. Only in the multi-arm STAMPEDE trial was there a docetaxel and an abiraterone arm, and a stronger effect in terms of time to treatment failure was reported with abiraterone than with docetaxel, but again these were not in a direct comparison. In contrast, there was no difference in overall survival or tumor-specific survival. There is also an indirect comparison for apalutamide to the docetaxel arm of the STAMPEDE study (see above); an IQWIG assessment of this comparison found evidence of a non-quantifiable added benefit for apalutamide due to advantages in the side effects category. Overall, the new generation of hormone drugs seems to have a more favorable side-effect profile; on the other hand, the duration of therapy is longer, and for high-risk patients taking abiraterone, the long-term administration of glucocorticoids should be considered as compared to docetaxel chemotherapy.

Overall, the choice of therapy should therefore be made for the appropriate indication, taking into account patient preferences, possible side effects, and the existing individual comorbidity profile after participatory decision making. It should be taken into account that relatively long survival times can also be achieved with a "simple" AD, such that, for example, a combination treatment does not necessarily have to be administered in cases of presumed limited life expectancy. In this sense, the restriction of the recommendations to patients with good performance status (ECOG 0-1) should also be understood, since it is important to doubt that patients who do not improve accordingly in performance status within 3 months on an initial hormone therapy will actually benefit from early combination treatment.

Re: Recommendation 7.25

a)     Immediate hormone-ablation therapy is associated with prolonged progression-free survival [734]. However, as stated in Chapter 6.7 "Primary Hormone Ablative Therapy and Watchful Waiting" in non-metastatic prostate cancer, the results are inconclusive in both non-metastatic and metastatic disease with respect to overall survival. However, due to the good response rates and prolongation of progression-free survival in the symptomatic metastatic stage, a strong recommendation is made for immediate hormone-ablation therapy. Cause-oriented treatment is clearly preferable to symptomatic treatment. In addition to prolonging progression-free survival, there is evidence that early initiaton of androgen deprivation reduces complications due to progression of the underlying disease (e.g., due to pathological fractures) [737].

With regard to both the indication and other aspects of androgen deprivation (AD), the situation of patients with localized prostate cancer cannot be reliably differentiated from that of patients with metastatic PCa on the basis of published analyses. Furthermore, there is no evidence that hormonal patients in localized tumor stages behave differently with respect to response to AD than those with metastatic PCa. Accordingly, both the meta-analysis by Wilt 2001 [734] and the ASCO guidelines of 2004 and 2007 [36][737], respectively, and this guideline, used study results from patients with localized and advanced stages for the recommendations [806].

b)     A similar recommendation can be found in the chapter Watchful Waiting and Hormone-Ablation Therapy Alone in Non-metastatic Prostate Cancer. The recommendation cites agents that have been effectively used for AD in randomized-controlled trials. The systematic review by Kunath et al. 2019 [789] includes studies of orchidectomy and LH-RH agonists. In addition, estrogens or DES have been used in the VACURG [739] studies. Iversen 2006 [728] uses bicalutamide; Studer 2006 [729] also uses LH-RH agonists or orchidectomy. The use of GnRH blockers has been derived from its effect on lowering testosterone levels, as seen in LH-RH agonists. Of the GnRH antagonists, Abarelix has been approved since 2005, and Degarelix since February 2007 for the indication of hormone-ablation therapy in advanced prostate cancer (Jakob et al. 2017/2021) [807]. Monotherapy with steroidal antiandrogens is associated with shorter progression-free survival compared with LH-RH-analog therapy and should not be recommended [36].

c)       Two meta-analyses [795][796] form the basis for this recommendation, each summarizing primary studies comparing continuous and intermittent AD. The majority of included studies, including the largest study with more than one thousand patients [797], had an inclusion criterion of a drop in PSA level below 4 ng/ml following an induction period (up to 7 months) to then be randomized to a continuous or intermittent treatment group. For patients with higher levels after the AD induction phase, the guideline group believes there are insufficient data on the efficacy and safety of intermittent AD, and therefore it is not recommended for this indication.

(d)     The available pooled studies, mostly of moderate quality, included patients of different stages and appropriate subgroup analyses were not performed. The meta-analyses, as well as the largest study, which included only metastatic stages, cannot demonstrate clear inferiority or superiority of any of the treatment options in terms of overall or cancer-specific survival, as well as time to progression. However, the Hussain trial showed a non-significant survival benefit of a median 5.8 vs. 5.1 years for continuous AD, with clear limitations. Data are also unclear or insufficient regarding the potential for harm from side effects as well as effects on quality of life; therefore, special consideration should be given to the individual patient's circumstances.

Re: Recommendations 7.26 - 7.28

Treatment for patients with advanced prostate cancer has changed fundamentally in recent years. In particular, the introduction of new drugs for systemic therapy has led to an increase in survival at different stages of the disease.

The treatment of castration-resistant prostate carcinoma is a palliative therapy. This fact must be taken into account when determining the indication for therapy. Patients should be informed accordingly.

The therapeutic options involve different disciplines. An interdisciplinary tumor conference for castration-resistant prostate cancer includes urology, internal oncology, pathology, radiotherapy and radiology, and depending on the symptoms and the therapeutic option to be discussed, also nuclear medicine, orthopedics, neurosurgery or other specialist disciplines.

The goals of therapy are to prolong survival, relieve symptoms, improve or maintain quality of life, and prevent complications. Therapy-associated side effects must be critically weighed up against the outcomes of therapy. Decision-making requires comprehensive patient education, which must include all aspects to be weighed up for a treatment decision.

Local treatment options for symptomatic or symptom-prone lesions relate to surgical and/or radiotherapeutic interventions. In addition to its use in the treatment of bone metastases, palliative radiotherapy is used for symptomatic locoregional pelvic tumors (primary tumor or lymphoma), especially for painful, obstructing, and/or bleeding tumors (can also be applied as re-radiation), in rare cases for pulmonary metastases (in the presence or threat of dyspnea), or as irradiation of cerebral metastases [808][809].

Systemic treatment options in patients with castration-resistant prostate cancer include antihormonal treatments, chemotherapy, and radionuclide therapy, as well as bone-modifying agents and supportive therapy measures.

The supporting literature is set out in the specific recommendations that follow. The order is based on clinical indications:

-       Non-metastatic castration-resistant prostate cancer (nmCRPC).

-       Metastatic castration-resistant prostate cancer (mCRPC).

o    Asymptomatic or mildly-symptomatic patients.

o    Symptomatic patients

o    Treatment sequence following initial therapy with at least one new hormonal agent (new hormonal agent)

o    Treatment sequence following docetaxel

o    Treatment sequence following androgen receptor-targeted treatment

Re: Statement 7.29

Chemotherapy is a therapy with relatively limited therapeutic breadth. The occurrence of adverse drug reactions (ADRs) is the rule, not the exception, even with standard-dose therapy. The toxicity of systemic therapy is classified using Common Terminology Criteria for Adverse Event (CTCAE)[http://evs.nci.nih.gov/ftp1/CTCAE/About.html]. In addition to this objective toxicity, the subjective burden on the patient plays a significant role. Both have to be recorded and weighted up in a structured way in the context of treatment decisions (continuation of therapy, dose modification and discontinuation of therapy). Therapeutic burdens are more likely to be accepted if the (supposed) advantages of therapy are great. Discontinuation of therapy, a high rate of toxicity, and dose modifications are therefore indications of low suitability for treatment.

Re: Statement 7.30

Given that aging is a highly heterogeneous process and conventional clinical history taking and physical examination often miss age-associated changes [810][811] and these changes are associated with outcome variables [812], in patients with cancer aged 70 years and older, performance of an assessment prior to initiation of chemotherapy is recommended by national and international professional societies [813][814][815][816]. Specific recommendations for patients with prostate cancer have been made by the International Society of Geriatric Oncology (SIOG) [418][817].

In the CRASH (Chemotherapy Risk Assessment for High Aged Patients) study, Extermann et al. analyzed variables associated with the occurrence of grade 3-4 nonhematological toxicities and grade 4 hematologic toxicities according to CTCAE criteria [818]. 518 patients aged 70 years and older treated with chemotherapy were included [818]. Using the following variables, 4 risk groups (low = 7 %; intermediate-low = 23 %; intermediate-high = 54 % and high = 100 %) could be created for the occurrence of grade 4 hematologic toxicities: Lymphocytes, ASAT, Instrumental Activities of Daily Living (IADL), LDH, diastolic blood pressure, and MAX-2 value of chemotherapy [818]. For the occurrence of grade 3-4 non-hematologic toxicities, 4 risk groups (low = 33 %; medium-low = 46 %; medium-high = 67 % and high = 93 %) could be identified using the variables: Hb, albumin, creatinine clearance, ECOG-PS, Mini-Mental Status Examination (MMSE), Mini-Nutritional Assessment (MNA), and MAX-2 value of chemotherapy form [818]. Docetaxel 75 mg/m² body-surface 3-weekly has a score of 2 on the CRASH score, while docetaxel 35 mg/m² body-surface weekly has a score of 0 on the CRASH score [818].

In a prospective cohort study that included 500 patients aged 65 years and older, median 73 years (range 65–91 years), Hurria et al. examined predictors of chemotherapy associated toxicity grade 3-5 [819]. Among others, geriatric assessment variables (mobility limitation, falls, need for assistance in instrumental activities of daily living (IADL), limitation of social activities) were predictors of low, intermediate, or high risk of grade 3-5 toxicities [819].

In a prospective cohort study that included patients aged 70 years and older who were scheduled to receive first-line chemotherapy, Soubeyran et al. analyzed geriatric assessment variables for their role in predicting death within the first six months [820]. Advanced disease, limitation in the Mini-Nutritional Assessment, and limitations in mobility were associated with 6-month mortality [820].

Re: Recommendation 7.31

The background to this recommendation is the fact that androgen-independent or castration-resistant disease can only be assumed once adequate castration-equivalent testosterone suppression has been implemented. Therefore, a serum testosterone level below 20-50 ng/dl is the target, and this is documented (definition of castration resistance based on the EAU guideline: serum testosterone <50 ng/dl with concomitant biochemical (three consecutive PSA increases one week apart yielding two increases of 50 % above baseline, and a PSA level >2 ng/mL)) or radiological progression [674][821]. In progressive disease, testosterone levels should be monitored. Progression of disease may be evident clinically, biochemically (see definition above), or on imaging.

In the STAMPEDE trial, a minority (18% of patients) received early systemic docetaxel treatment in addition to radiotherapy for the primary tumor [890]. If docetaxel treatment is given, sequential therapy is recommended; by contrast, in the STAMPEDE trial androgen deprivation was given "lifelong", i.e., also simultaneously with radiotherapy [890].

In addition to chemohormone therapy, abiraterone (high risk only) and apalutamide, and possibly enzalutamide in the future, are currently approved for hormone-sensitive metastatic prostate cancer. It is therefore possible that in the future the new hormone preparations mentioned above will also play a role in this indication in addition to radiotherapy of the prostate (or radical prostatectomy, see below) or in addition to metastasis-directed therapy (cf. 7.3).

The oncological efficacy of metastasis-directed treatment in oligometastatic prostate cancer has not been adequately demonstrated to date. This is especially true for survival endpoints. Delay of androgen deprivation or tumor progression is possible with metastasis-directed treatment [893][895]. More studies, particularly randomized trials, are needed to determine whether these effects favorably affect disease progression or prolong survival [895]. In the ORIOLE study, the impact of metastasis-directed ablative radiation in patients with (metachronous) oligometastatic hormone-sensitive PCA was investigated in a randomized prospective phase-II setting. The initial diagnosis did not yet require PSMA imaging, so that, according to protocol, lesions that could only be visualized on optional PSMA PET may not have been treated. Metastasis-directed treatment significantly prolonged PSA progression in the entire intervention group, and higher-grade adverse events(≥ G3) were not observed. In addition, a subgroup analysis of a cohort in which all PSMA positive lesions received ablative radiation also showed longer metastasis-free survival [230].

However, in the case of metastasis-directed treatment for oligometastatic disease, it seems reasonable to additionally educate patients about the possibility of intensified systemic therapy. Until proven otherwise, it can be assumed that metastasis-directed treatment in oligometastatic prostate cancer cannot permanently replace intensified systemic therapy (i.e., with abiraterone, apalutamide, docetaxel, or enzalutamide).

Re: Recommendation 7.58

Bone metastases are present in over 80% of patients with metastatic prostate cancer [903], [904]. Typical predilection sites are the spine, pelvis, and chest [905]. For prostate cancer, osteoblastic metastases are predominant. The median survival time for known bone metastases is approximately three years [906]. Symptomatic complications predominantly include pain, hypercalcemia, pathologic fractures, and spinal cord compression in vertebral fractures.

Randomized controlled trials for the prevention of skeletal-related events in bone metastases are available for the mentioned treatment modalities (percutaneous radiotherapy, systemic application of radionuclides) [868], [896], [897]. The indications are explained further in the following recommendations. Side effects and contraindications need to be considered. Based on the available data, it must currently be concluded that a prolongation of life is generally not achieved with a specific treatment of bone metastases, even with treatment of singular metastases. Individual studies on the benefit of percutaneous radiotherapy or the administration of radionuclides, in which positive survival time effects were achieved are contrasted with other studies in which such effects were not demonstrable [897].

The standard of medicinal pain treatment for tumor patients is the WHO staging scheme. In summary, this regimen provides for a non-opioid analgesic and/or adjuvants in stage I, a weak-acting opioid (not under controlled substances law) combined with a non-opioid analgesic, coanalgesics and/or adjuvants in stage II, and a strong-acting opioid (controlled substance) combined with a non-opioid analgesic, coanalgesics and/or adjuvants in stage III [907].

Re: Recommendation 7.59

Local percutaneous radiotherapy of painful singular bone metastases is a palliative measure. The available randomized controlled trials show partial or complete pain relief after approximately three to eight days in more than 80% of patients. In at least 50% of patients, this pain relief lasts six months or longer. Assuming a single application of at least 6 Gy (usually 8 Gy), the degree and duration of pain relief in the studies do not correlate significantly with the radiation regimen used (single dose versus fractionated administration) [896], [908], [909]. However, the data show a significantly increased rate of subsequent retreatment after single-dose radiation compared with a fractionated regimen [896], [910].

Acute adverse effects of radiation therapy include nausea and vomiting in approximately one quarter of cases [897]. The incidence of spinal cord compression or pathologic fractures around the target volume are low later in the course. The incidence is 1-2% for spinal stenosis, and 1-4% for pathologic fractures in randomized controlled trials [896]. A twofold to threefold increased incidence of pathologic fractures compared with fractionated radiotherapy was reported for single-dose radiation when studies up until 2001 were analyzed [910]. A more recent meta-analysis with seven additional studies through July 2006 found no significant differences with respect to pathologic fractures [911]. Remineralization after six months is significantly better in patients receiving fractional radiotherapy compared to patients receiving single-dose radiotherapy (data from one study, [912]).

Data from one retrospective and four prospective case series (with a total of 545 patients) are available on the effect of radiotherapy in tumor-related spinal stenosis or spinal cord compression [896]. Signposts for a possible success of the therapy are the early diagnosis and the radiosensitivity of the tumor. The studies show an improvement of the presenting neurological symptoms in about 25-64% of patients ([896], but the definitions for this are heterogeneous.

For comparison of primary combined therapy with surgical decompression and percutaneous radiotherapy (10 x 3 Gy) vs. radiotherapy alone, a 2005 RCT is available [913]. A Cochrane review on the topic did not identify any other studies on surgical decompression combined with radiotherapy vs. radiotherapy alone until 2008 [914]. Searches through 3/2011 also yielded no new RCT on the topic. The RCT showed significantly better outcomes in patients (n = 103) with proven spinal canal compression and neurological symptoms < 48 hours in terms of maintaining or regaining walking ability with combined therapy (84% vs. 57% and 62% vs. 19%, respectively). The interventions were each performed within 24 hours. Patients whose symptomatology was due to unilocular compressive metastasis of a primary nonhigh-radiation tumor were included. Surgical intervention after unsuccessful radiotherapy showed far worse results than primary combined treatment in this RCT. Because of the significant result in favor of combination therapy, the RCT was discontinued after interim analysis. Among the patients included were 19 patients with prostate cancer. No differentiation is made whether this was hormone-naïve or castration-resistant.

In hormone-naïve patients with symptoms of spinal compression without pathologic fracture, there is also the possibility of (emergency) hormone-deprivation therapy.

The decision for surgical intervention should be made on an individual basis. Surgical stabilization should usually be considered in cases of impending fracture risk or after pathologic fracture, depending on the location. The indication should be determined on an interdisciplinary basis with the involvement of a specialist in radiological diagnostics, a radiation therapist, and a representative of a surgical specialty (neurosurgeon, orthopedist, or trauma surgeon). This is often an emergency situation, most of which requires immediate steroid medication.

Postoperative local radiotherapy (10 x 3 Gy) significantly reduces the rate of reoperation due to prosthesis loosening [897].

Re: Recommendation 7.60

In randomized controlled trials radionuclides have been shown to be effective in reducing pain in multiple diffuse bone metastases in the hormone refractory stage.

The radionuclides approved for the indication are strontium-89 (St-89), samarium-153 (Sm-153), and rhenium-186 (Re-186). The radionuclides Sr-89 at a dosage of 150-200 MBq and Sm-153 at a dosage of 1.0 mCi/kg have been best studied for intervention (several phase III RCTs available in each case). For Sr-89 application, pain relief sets in slightly later, at two to three weeks after initiation of treatment, than for Sm-153 [896] application, at one to two weeks. The occurrence of fewer new pain foci has been described for Sr-89 compared to Sm-153. A phase-III RCT is available for the efficacy of Re-186.

For the radionuclides Re-188 and Sn-117, only phase-II studies are available so far, but no phase-III studies.

The pain relief achieved in 60-80% lasts for a median of two to four months in the majority of patients. In addition, fewer new painful foci are reported compared to placebo [896], [897]. When determining the indication, the expected side effects should be considered. Nausea and/or vomiting occurs only in a minority of patients (up to 10%) [897]. However, treatment regularly results in mild to moderate myleosuppression ([896], [897]. Recovery of leukocyte counts to baseline values is observed with Sm-153 after approximately eight weeks [915]. This time period can also be expected with application of Re-186, whereas leukocytes do not return to baseline values after application of Sr-89 until after approximately twelve weeks. Patients should also be informed about a possible temporary increase in symptoms (flare phenomenon) before definitive pain relief (GCP).

Data on radium-223 is discussed in Recommendation 7.49 [851].

Re: Recommendation 7.61, Recommendation 7.62, and Recommendation 7.63

With regard to the treatment of bone metastases to prevent or delay symptomatic skeletal-related events, the guideline group decided to recommend both the RANKL antibody denosumab and the bisphosphonate zoledronic acid. This recommendation applies only to castration-resistant patients. By contrast, the administration of osteoprotective agents is not recommended in hormone-naïve patients with bone metastases: The majority of the identified studies show no advantage of using zoledronic acid in hormone-naïve patients with regard to time to skeletal-related events (SRE) and survival; no studies on other bisphosphonates and denosumab could be identified for this patient group. In addition to uncertain or unproven benefits, the agents are associated with potential risks: renal dysfunction (zoledronic acid [916]) or osteonecrosis of the jaw and hypocalcemia (denosumab [917], [918] (see also Recommendation 7.52), On this basis, the guideline authors advise against the use of osteoprotective agents in hormone-naïve patients with bone metastases. If treatment is indicated due to osteoporosis, this remains excluded from the negative recommendation. As generally recommended in the S3 guideline on supportive therapy (http://leitlinienprogramm-onkologie.de/Supportive-Therapie.95.0.html) [919], at least 500 mg calcium and 400 I.U. vitamin D should be given orally daily with both agents. The evidence for the bisphophonates or for the RANKL antibody in bone metastasis complications is presented below. No data are available on a "switch" of ongoing therapy with zoledronic acid to therapy with denosumab.

Evidence on bisphosphonates

In the combined analysis, the available studies show a borderline-significant effect with the use of different bisphosphonates with regard to the occurrence of skeletal-related events in bone metastases of castration-refractory prostate carcinomas (odds ratio: 0.79; 95% CI: 0.62-1.0; p = 0.05). Existing studies evaluating for this question used clodronate, pamidronate, and zoledronate [868]. Most bisphosphonates are effective only in bone metastases with osteolytic components; this explains the partial lack of efficacy in osteoblastic metastatic prostate carcinoma. For more information on the associated risks, see Recommendation 7.52 on osteonecrosis of the jaw.

In the single analysis, zoledronic acid in combination with vitamin D and calcium was the only substance studied to show a significant reduction for the occurrence of skeletal-related events compared to placebo administration (+ vitamin D and calcium), especially for pathologic fractures [867], [906], [920]. Therefore, the recommendation on zoledronic acid as a bisphosphonate was changed to a strong recommendation in the second guideline update in 2014. It should be noted that the evidence of the effect of zoledronic acid is based on a single study (n = 634). The absolute difference compared with placebo for the occurrence of skeletal-related events was 11% (44.2% in the placebo group, 33.2% in the zoledronic acid group). The time to occurrence of skeletal-related events was several months longer, this was statistically significant (placebo median 321 days, zoledronic acid median not yet reached after 15 months; p = 0.01). An update search on the third update identified four additional RCTs (three open label, one blinded) [794], [899], [900], [901]. Three studies find no significant effects compared to placebo in terms of time interval to skeletal-related events (SRE). The smallest study (n=188) [899] reports a significant prolongation of the SRE-free interval by 2.4 months (from 11.2 to 13.6 months, but the included patients in this study were already castration-resistant. Other endpoints including progression-free and overall survival again did not differ between zoledronic acid and placebo groups. Bisphosphonates including zoledronic acid are not approved for the indication of pain management, and their use constitutes an off-label use.

Evidence on denosumab

The human monoclonal antibody denosumab targets RANKL, thereby inhibiting osteoclast-mediated bone destruction.

Fizazi et al. published the results of a multicenter non-inferiority RCT (n = 1,904) in spring 2011 [866]. The primary endpoint of this study was to compare denosumab with zoledronic acid in terms of time to first skeletal-related event (defined as pathologic fracture, percutaneous osseous radiotherapy, surgical osseous intervention, or spinal cord compression) in patients with castration-resistant prostate cancer. Patients with creatinine clearance of < 0.5mL/s were not included because zoledronic acid cannot then be applied. In contrast to zoledronic acid (i.v.), denosumab is administered subcutaneously and can also be used in patients with impaired renal function. With denosumab, the first skeletal-related event occurred a median of 3.6 months later than with zoledronic acid (20.7 months versus 17.1 months), with an overall median survival of 22 months for both groups. The analysis was based on intention to treat. The difference was statistically significant in favor of denosumab in terms of hazard ratio – in terms of noninferiority, but also when tested for superiority. However, the confidence intervals of the absolute results overlapped in months. For denosumab (median 17.1 months), the lower limit of the 95% confidence interval is 18.8 months (upper limit 24.9 months); for zoledronic acid, the upper limit of the 95% confidence interval is 19.4 months (lower limit 15 months). This means that based on the lower and upper limits given in the confidence intervals for the "true value" there is uncertainty regarding superiority of denosumab. For this reason, the result is estimated that the equivalence of denosumab with zoledronic acid in terms of time to skeletal-related events can be confidently inferred from the study results, but not superiority.

The difference in absolute events over the entire observation period is 5% (36% denosumab, 41% zoledronic acid). The other study identified with the update search, [898], with the same study design, also reports an absolute risk reduction for skeletal-related events of 4.7% (35.9% denosumab, 40.6% zoledronic acid). It should be noted in this study that it has significant methodological flaws (patient characteristics not reported, statistics not transparent) and was supported by the manufacturer of the monoclonal antibody during manuscript preparation.

With regard to individual adverse effects, there were statistically significantly more hypocalcemias under denosumab (13% denosumab, 6% zoledronic acid). Significantly higher were grade 3+4 adverse events overall (72% vs. 66%) using the CTCAE (Common Terminology Criteria for Adverse Events version 3.0) classification. A Red Hand Letter from the manufacturer [907] reported severe symptomatic hypocalcemia, including cases with fatal outcomes, in patients treated with denosumab. To minimize the risk, monitoring of calcium levels was recommended by the manufacturing company in patients with severe renal impairment (creatinine clearance < 30ml/min) or in patients requiring dialysis. The S3 guideline on supportive therapy (http://leitlinienprogramm-onkologie.de/Supportive-Therapie.95.0.html) [919] recommends the simultaneous administration of at least 500 mg of oral calcium and 400 I.U. vitamin D daily as a preventive measure.

Under denosumab 2% and under zoledronic acid 1% of osteonecrosis of the jaw occurred, and secondary malignancies occurred just as frequently in each case. The differences are not statistically significant. The relatively short follow-up period should be considered. The risk of osteonecrosis of the jaw increases with longer treatment or follow-up duration (see Recommendation 7.51). For prevention of osteonecrosis of the jaw, please refer to the corresponding S3 guideline (http://www.awmf.org/leitlinien/detail/ll/007-091.html).