Treatment of Advanced Prostate Cancer: Where Are We in 2019?

Until 2010, docetaxel (taxane) was the only agent with OS improvement and it was the standard firstline therapy for mCRPC. The phase III study TAX 327 randomized 1006 mCRPC patients to receive Docetaxel 75mg/m² every 3 weeks for up to 10 cycles versus Docetaxel 30mg/m² weekly versus Mitoxantrone 12mg/m² every 3 weeks and showed an improvement of OS for Docetaxel every 3 weeks (19.2 vs 17.8 vs 16.9 months, HR = 0.79, p = 0.004) in all subgroups. (symptomatic vs asymptomatic, presence vs absence of visceral metastases, good KPS vs poor KPS and age < 68 years vs = 69 years), The patients in the docetaxel every 3 weeks group also presented a higher PSA response rate of 50% and improvement in QoL.(32 40) Adverse events (AE) such as neuropathy (30%), fatigue (53%), alopecia (65%), diarrhea (32%) and neutropenia grade 3 or 4 (30%) were more common with docetaxel, however the treatment discontinuation rate was similar between groups.

Another option for the use of docetaxel is the schedule every 2 weeks, based on a randomized phase 3 study of 177 patients which compared docetaxel 50mg/m² every 2 weeks versus docetaxel 75mg/m² every 3 weeks and demonstrated an increase in time to treatment failure (5.6 vs 4.9 months, p = 0.01), a lower incidence of grade 3 and 4 neutropenia (36 vs 53%) and lower incidence of febrile neutropenia (4 vs 14%) for the use every 2 weeks.(41)

Cabazitaxel is a new generation taxane developed to overcome resistance to docetaxel. In addition, this drug has the capacity to cross the blood-brain barrier.(42) Its activity was demonstrated in the phase 3 TROPIC trial that randomized 755 mCRPC patients who had progression during or after therapy with docetaxel. These patients received cabazitaxel 25 mg/m² every 3 weeks associated with prednisone 10 mg daily for up to 10 cycles (C+P) or mitoxantrone 12 mg/m² every 3 weeks also associated with prednisone.

With a median follow-up of 13.7 months, this study showed an improvement in OS (15.1 vs 12.7 months, HR = 0.70, 95% CI 0.59-0.83, p < 0.0001) and in progression-free survival (PFS) (2.8 vs 1.4 months, HR = 0.74, 95% CI 0.64-0.86, p = 0.0002) for those treated with C+P.(33) Treatment with C+P was associated with higher toxicity, such as diarrhea (47%), fatigue (37%), neutropenia grade 3 and 4 (82%), and febrile neutropenia (8%).

Due mainly to hematological toxicity, the non-inferiority phase 3 PROSELICA trial evaluated the efficacy of cabazitaxel in a lower dose. In this study, 1200 patients with mCRPC who failed docetaxel were randomized to cabazitaxel 25mg/m² (C25) every 3 weeks or to cabazitaxel 20mg/m² (C20) every 3 weeks and no inferiority was demonstrated in OS or PFS at lower dose. In addition, grade 3 and 4 AE (39.5% C20 vs 54.5% C25), grade 4 neutropenia (21.3% C20 vs 48.6% C25) and neutropenic infection (2.2% C20 vs 6.1% C25) were less frequent at the lower dose.(43)

Therefore, cabazitaxel at a dose of 20mg/m² every 3 weeks is one of the options for second-line mCPRC (after docetaxel failure) treatment.

Abiraterone acetate is a hormonal agent that inhibits androgen biosynthesis through inactivation of the CYP-17 enzyme, acting in the three sources of androgen production: testis, adrenal gland and intra-tumor. Its activity was demonstrated in mCRPC as first line treatment and as second line (after docetaxel failure).

The phase 3 COU-AA 302 trial randomized 1088 chemotherapy-naïve mCRPC patients, asymptomatic or oligosymptomatic, to abiraterone 1g/day and prednisone 10mg/day or to placebo and prednisone 10mg/ day. A follow-up of 49 months showed improvement of OS (34.7 vs 30.3 months, HR = 0.81, p = 0.003) and rPFS (16.5 vs 8.2 months, HR = 0.52, p < 0.0001) for the abiraterone group. However, 44% of patients in the placebo group received abiraterone later, which may have influenced the magnitude of the results.(44) All secondary outcomes (time to initiation of chemotherapy, skeletal events, progression of PSA, deterioration of KPS, etc.) and QoL were significantly favorable for abiraterone.(45)

The COU-AA 301 phase 3 trial randomized 1195 mCRPC patients who failed to docetaxel to abiraterone 1g/day and prednisone 10g/day or placebo and prednisone 10mg/day. With a 20 months median follow-up an improvement in OS was demonstrated (15.8 vs 11.2 months, HR = 0.74, p < 0.001) for the abiraterone group. This group had also benefit in all secondary outcomes (time to PSA progression, rPFS, PSA response).(46) An exploratory analysis of this study also showed benefit of abiraterone for pain relief, time for pain worsening and reduction of skeletal-related events.(47)

In general, toxicity of abiraterone is related to the excess of mineralocorticoids, so the concomitant use of prednisone is necessary. Water retention (33%), hypertension (11%) and hypokalemia (18%) are among the most frequent AE. Other AE observed are changes in liver function tests (11%) and miscellaneous cardiac disorders (16%). Despite the toxicity mentioned above, the rate of discontinuation of treatment related to AE was similar to that of the placebo group.(35)

Thus, abiraterone 1g/day and prednisone 10mg/day is an option for mCRPC either in the first or second line treatment (after docetaxel).

Enzalutamide is a strong AR blocker, thus, it acts directly by blocking the AR, preventing nuclear translocation of the AR and inhibiting the binding of the receptor to the nucleus DNA. Its efficacy in mCRPC was demonstrated in M0 and M1 disease setting, in both chemotherapy-naïve patients and in those who failed to docetaxel. Three important phase 3 trials evaluated enzalutamide in mCRPC setting. Two trials evaluated enzalutamide in metastatic castration-resistant prostate cancer patients (M1CRPC) and one trial included only non-metastatic castration-resistant prostate cancer patients (M0CPRC) with highrisk disease (rapidly rising PSA).

The PREVAIL trial randomized 1717 chemotherapy-naïve mCRPC patients, asymptomatic or oligosymptomatic, to enzalutamide 160mg/day or placebo. With a median follow-up of 22 months it was shown improvement in OS (34.7 vs 30.2 months, HR = 0.71, p < 0.001) and in secondary outcomes (rPFS and reduction of skeletal events) for the enzalutamide group. In addition, there was also a significant improvement in QoL for the enzalutamide group.(36 48) Another study, the AFFIRM trial randomized 1199 mCRPC patients previously treated with docetaxel to enzalutamide 160mg/day or placebo. The enzalutamide group showed improvement in OS (18.4 vs 13.6, HR = 0.63, p < 0.001) and in all secondary outcomes (PSA response rate, time to PSA progression, rPFS). In a secondary analysis of the trial, those treated with enzalutamide had a significant improvement in QoL.(37 49)

The most common AE related to enzalutamide are fatigue (34%), hot flashes (20%) and diarrhea (21%). In the AFFIRM trial some patients that received enzalutamide presented seizures and, due to this, those with a past of seizures and/or recent brain event were excluded from the PREVAIL trial.(36)

Thus, enzalutamide 160 mg/day is an option for M0CRPC high-risk disease treatment and for M1CRPC first and second line treatment.

Sipuleucel-T is an autologous vaccine of dendritic cells developed to increase T cell-mediated immune response against prostatic acid phosphatase. Its activity in mCRPC was demonstrated in the phase 3 IMPACT trial. This trial randomized 512 mCRPC patients (asymptomatic or oligosymptomatic, previously treated with docetaxel or not) to sipuleucel-T every 2 weeks for 3 cycles or placebo. With a median follow-up of 34 months it was showed an OS improvement (25.8 vs 21.7 months, HR = 0.78, p = 0.03) for the vaccine group. However, neither improvement in PFS nor in PSA response rate was observed.(38)

This therapy is well tolerated and, as in other immunotherapies, the most common vaccine-related AE were chills (53%), fatigue (41%), and fever (35%).(50)

Therefore, sipuleucel-T is an option of for mCRPC, but it is not available outside the USA.

Radium-223 is an a-particle emitter therapeutic agent which has tropism for bone tissue. It is able to achieve bone metastases with high energy in a small radius of action, minimizing toxicity in the bone marrow.

The phase 3 ALSYMPCA trial randomized 921 mCRPC patients previously treated or not with docetaxel, symptomatic and with bone metastases only (no known visceral metastases) to receive radium-223 every 4 weeks up to six cycles or placebo. This trial demonstrated an improvement in OS for those treated with radium-223 as first or second line of treatment (14.9 vs 11.3 months, HR = 0.70, p < 0.001). Treatment was well tolerated and there was no significant difference in grade 3 or 4 toxicity when compared with the control group.(39 51) However, the reduction to the first bone event was only observed in the subgroup of patients who received radium-223 as second line of treatment.(95) Regardless of the line of treatment, the use of radium-223 was associated with a significant improvement in QoL.(96)

Thus, radium-223 is another option of treatment - as first or second line (after docetaxel) - for mCRPC with bone metastases.

The options for first-line mCRPC treatment available in Brazil, the inclusion criteria, as well as the most common adverse effects are summarized inTable 3.

Unlike TAX 327 and ALSYMPCA trials, in which 45% and 100% of the patients, respectively, were symptomatic, in PREVAIL and COU-AA 302 trials the symptomatic patients were excluded. Patients with visceral metastasis were excluded of COU-AA 302 and ALSYMPCA trials, but represented 11% of PREVAIL trial patients and 22% of TAX 327 trial patients.(32 34 36 39) Regarding to lymph node disease, those with lymph node enlargement (> 3cm) were excluded from the ALSYMPCA trial.

Recently, a prospective multicenter phase 3b trial of 839 mCRPC patients evaluated Radium-223 in some different scenarios of the ALSYMPCA trial. In this trial, 20% of the patients were asymptomatic and Abiraterone (18%) or Enzalutamide (5%) were allowed to be used associated with Radium-223. Subgroup analysis according to the symptomatology and level of alkaline phosphatase demonstrated a greater OS for asymptomatic patients and for those with alkaline phosphatase levels within normality.(95)

The improvement of OS in these subgroups raises the question whether this benefit occurs from the early onset of Radium-223 (at the earliest stage of the metastatic disease) or from the more indolent and less aggressive disease.

As previously discussed, due to seizures in some patients that received enzalutamide after prior therapy with docetaxel (AFFIRM trial), those with a historic of seizures or some recent brain event (e.g. stroke) were also excluded from the PREVAIL trial.(36 37)

Recent data from patients in a phase 2 trial evaluating the sequence of first-line treatment with abiraterone and enzalutamide in mCRPC showed a worsening of depressive symptoms and an increase of cognitive impairment in those treated with enzalutamide.(66)

Unlike docetaxel and abiraterone, the use of enzalutamide and radium-223 does not require association with corticosteroids. Some retrospective series suggest that there is no negative impact on the efficacy and safety of abiraterone when used with a lower dose of corticosteroid (prednisone 5 mg/day), as well as the lower incidence of side effects related to prolonged corticoid use.(67 68)

Recently docetaxel and cabazitaxel were compared in first-line treatment. The FIRSTANA trial randomized 1168 mCRPC patients to receive cabazitaxel 20mg/m² or cabazitaxel 25mg/m² or docetaxel 75mg/m² as first line treatment. Despite the higher response rate with Cabazitaxel 25mg/m², there was no difference in OS (24.5 vs 25.2 vs 24.3 months, p = 0.99). The AE and the discontinuation rate of the treatments were similar.(69)

Currently there is another possible setting for the second-line treatment: those patients who received chemohormonal as first-line treatment in the mCSPC setting. There was no standard treatment for those patients.Table 4 summarizes the alternatives that are offered after progression with initial chemohormonal treatment.

A retrospective analysis of the GETUG-AFU 15 trial demonstrated a PSA response rate of 13-70% and symptom improvement in 16-33% of patients when exposed to other agents following progression.(70) Exposing these patients to docetaxel again does not seem to be a good option, considering that docetaxel re-treatment in mCRPC, despite the PSA response rate, did not show improvement in OS.(71) In this setting, the most appropriate would be the agents that improve OS in the second-line treatment after docetaxel (Abiraterone, Enzalutamide or Cabazitaxel - see details below).

The second line treatment options available in Brazil, some details of each trials and the most common AE associated with these therapies are specified inTable 5.

To this moment there are no studies comparing these second line therapies or evaluating their use in sequence. Therefore, the data and results of the main phase 3 trials and some retrospective analysis of the action of these drugs when used in sequence should be of great help when choosing the most appropriate treatment.

Considering the control arms of each clinical trial in second line treatment after failure to docetaxel, placebo plus prednisone and placebo alone where used as controls arms in the COU-AA-301 and AFFIRM trials respectively. In the other hand, the TROPIC trial control group received mitoxantrone, a chemotherapy with known palliative action in pain control and PSA response rate, however with no improvement in OS.(33) That could explain the results found in the update of the trial that showed no difference in pain palliation and QoL in comparison with cabazitaxel.(72)

The secondary endpoints of the trials were different. While COU-AA-301 and AFFIRM trials considered disease progression as evidence of tumor progression radiologically, the TROPIC trial defined disease progression as rise in PSA, evidence of tumor progression radiologically or worsening of the pain. Thus, treatment in the TROPIC trial may have been discontinued sooner in comparison to treatment in the COU-AA-301 and AFFIRM trial, therefore resulting in a shorter duration of treatment.

The TROPIC, COU-AA-301 and AFFIRM trial enrolled patients that progressed after treatment with docetaxel. In this context, 63% of the patients enrolled in the TROPIC trial and 45% of the patients in the COU-AA-301 had discontinued the treatment with docetaxel due to progression.(73 74) That may indicate that patients enrolled in the TROPIC trial, probably, presented a higher resistance to docetaxel.

At the same time that in the COU-AA-301 and AFFIRM trials the OS benefit was seen in patients regardless of age (= 65 vs < 65yrs), in the TROPIC trial there was no difference in OS in patients under 65 yrs. old.(33)

All secondary endpoints were positive in the three trials. Time to the first skeletal event was longer in the COU-AA-301 and AFFIRM trial, however this endpoint was not evaluated in the TROPIC trial.(33 35 37)

The median duration of treatment was 8 and 8.3 months for the COU-AA-301 and AFFIRM trials, respectively. Patients in the TROPIC trial received a median of six treatment cycles. The cross-over was allowed in the COU-AA-301 and AFFIRM trials and not allowed in the TROPIC trial.(33 35 37)

The toxicity related to each treatment has been discussed previously. It is important to emphasize that, being chemotherapy, the hematological toxicity observed in the TROPIC trial may be a limiting factor for treatment with cabazitaxel. Regarding enzalutamide, as previously discussed, historic of seizures could be a contraindication.(36 37) In the case of abiraterone, there is the possibility of hepatotoxicity with elevation of transaminases, which may be a limiting factor to its use in those with some type of underlying liver disease.

There is no good level evidence for the best sequence treatment to mCRPC in this setting. Retrospective data suggest lower response rates when abiraterone and enzalutamide are used in sequence, indicating perhaps the existence of a cross-resistance mechanism between them(81 82 85 86) (Table 6). Evidence also suggests that taxane chemotherapy may reverse or downregulate some of these resistance pathways, such as elimination of AR splice variants.(55) However, a retrospective series with 80 patients who were treated with abiraterone followed by enzalutamide versus abiraterone followed by docetaxel and then enzalutamide or enzalutamide followed by docetaxel and then abiraterone, demonstrated worse progression- free survival of PSA (4.4 vs 1.6 months, p < 0.01) and worse rPFS (5.7 vs 2.2 months) for the use of docetaxel intercalated with the new hormonal agents.(94)

Recently, exploring the sequencing of these two agents, a phase 2 trial with 202 M1CRPC patients evaluated abiraterone plus prednisone followed by enzalutamide (arm A) versus enzalutamide followed by abiraterone plus prednisone (arm B), both at PSA progression. The primary endpoints, PSA decline > 50% on second line therapy and time to second PSA progression, were better with arm A sequence (34 vs 4% and 2.7 vs 1.3 months, respectively).(83) In addition to this data, there are only retrospective series evaluating the activity of sequencing these agents in this setting (Table 6).

Although docetaxel was the main second line treatment in patients after disease progression on COUAA-302 (44% of the abiraterone and 58% of the placebo group receive some second line treatment)(34) and PREVAIL trial (40% of enzalutamide and 70% of placebo group receive some second line treatment),(36) the activity of chemotherapy (docetaxel or cabazitaxel) in this setting is uncertain.

Retrospective data from patients treated with docetaxel after abiraterone demonstrated an OS of 12.5 months, lower than the 19.2 months showed by the TAX 327 trial, which might suggest the hypothesis of some degree of cross-resistance between the agents.(77) On the other hand, in a post-hoc analysis of COU-AA-302 evaluating docetaxel activity in patients treated previously with abiraterone, a similar PSA response rate to the TAX 327 study was demonstrated.(84) While some series demonstrate no benefit with docetaxel in patients that failed abiraterone (primary resistance), others show responses in these patients and do not establish a direct relation between responders and non-responders to abiraterone and responders and non-responders to docetaxel.(77 78 79 80)

Given the limitations of these retrospective data, it can also be inferred that docetaxel maintains its activity after previous use of abiraterone.

However, the response or non-response to previous treatment with abiraterone or enzalutamide does not seem to be a predictive factor for response to cabazitaxel.(87) As well as cabazitaxel also has activity in those patients that failed (progression = 3 months) docetaxel.(84)

Although a considerable number of patients have good clinical conditions to receive some treatment in this setting, there is no standard recommendation. In the absence of randomized clinical trials, some small retrospective series summarized inTable 7 have demonstrated a PSA response rate ranging from 3 to 39% and a median OS of 7 to 16 months.(85 86 87 88 89 90 91 92 93)

Given the proven efficacy of immunotherapy in several solid tumors (lung and kidney cancer, melanoma, etc.), this treatment strategy has also been evaluated in mCRPC. Ipilimumab, a humanized anti-CTL-4 monoclonal antibody, was evaluated in two phase 3 trials. In the first trial 799 patients with mCRPC and at least one bone metastasis who had failed to docetaxel were randomized to receive bone-directed radiotherapy followed by either ipilimumab 10mg/kg or placebo every 3 weeks up to four doses. After median follow-up of 9.9 months there was no difference in OS (11.2 vs 10 months, HR = 0.85, 95% CI: 0.72-1.00, p = 0.053).(100) The other trial 602 asymptomatic or minimally symptomatic patients with chemotherapy-naive mCRPC were randomized to ipilimumab 10mg/kg or placebo every 3 weeks up to four doses. No significant difference was observed in primary endpoint OS (28.7 vs 29.7 months, HR = 1.11, 95.87% CI: 0.88-1.39, p = 0.36), but the secondary end point PFS was longer with ipilimumab (5.6 vs 3.8 months, HR = 0.67, 95.87% CI: 0.55-0.81).(101)

Pembrolizumab, a highly selective anti PD-1 humanized monoclonal antibody, demonstrated durable responses in 23 patients with mCRPC whose tumors have PD-L1 expression = 1 percent in a non-randomized phase 1b trial.(102) The KEYNOTE-199, a recent phase 2 trial, with 258 mCRPC who had failed at least a one novel endocrine therapy (abiraterone or enzalutamide) and chemotherapy evaluated pembrolizumab 200mg every 3 weeks and showed an objective response rate (primary endpoint) of 6 percent in those with PD-L1 expression positive. The secondary endpoint disease control rate at six months was 15 percent. Interestingly, patients with BRCA 1/2 or ATM mutations had better overall response rate.(103)

The activity combining immunotherapy was evaluated in a recent phase 2 trial with 78 asymptomatic or minimally symptomatic mCRPC patients who progressed after new generation hormone therapy and not received chemotherapy (cohort 1) or had received chemotherapy (cohort). The combination was nivolumab 1mg/Kg + ipilimumab 3mg/Kg every 3 weeks for 4 doses then nivolumab 480mg every 4 weeks. The primary endpoint ORR was 26 and 10% in cohort 1 e 2, respectively. In both cohorts the ORR was higher in patients with PD-L1 = 1% and DNA damage repair (DDR).(104)

It has been know that mCRPC can harbor some DNA repair gene mutation like BRCA1 or 2, ATM, CHEK2, NBN and PALB2 that may be associated with more aggressive disease and poorer OS.(105 106) Moreover, there are some trials showing durable antitumor activity in these mCRPC patients with treatment with poly-adenosine diphosphate ribose polymerase (PARP) inhibitors.(107) In this context, a phase 2 single-arm trial with 49 patients evaluated olaparib 400mg twice a day until disease progression, unacceptable toxicity or death. It should be noted that these mCRPC patients had failed to at least two prior treatments and all of them failed to chemotherapy with docetaxel. All patients had tissue evaluated for DNA gene repair mutations and this was detected in 33 percent of patients. Olaparib was associated with a higher response rate (primary endpoint) in those patients with DNA gene repair mutations when compared to those without mutations (88 vs 6%). Also rPFS and OS were longer in those with mutations (9.8 vs 2.7 months and 13.8 vs 7.5 months, respectively).(108)

Another phase 2 trial randomized 142 mCRPC patients who have failed to prior chemotherapy to olaparib and abiraterone or abiraterone. With a median follow-up of 15.9 months the rPFS was higher for olaparibe and abiraterone (13.8 vs 8.2 months, HR = 0.65, 95% CI:0.44-0.97, p = 0.034). In contrast, more AE were observed in combination treatment arm.(109)

Recently the activity combining Pembrolizumab and Olaparib in heavily treated mCRPC was evaluated in one cohort of KEYNOTE-365 phase 1b/2 trial with 40 patients that showed PSA response rate and ORR of 13 and 7%, respectively.(110)