De Novo Metastatic Prostate Cancer: Are We Moving towards a Personalized Treatment?

1. Introduction

According to GLOBOCAN 2020, almost a million and half of new cases of prostate cancer (PC) and approximately 400.000 PC-related deaths were reported in 2020 globally [1]. De novo metastatic hormone-sensitive PC (mHSPC) accounts for 5-10% of all PC diagnoses but it is responsible for nearly 50% of PC related deaths [2,3]. The incidence of de novo mHSPC is rising in the Western countries, probably due to the introduction of new diagnostic tool in the imaging of PC, such as PSMA-PET, and a reduction in PSA opportunistic screening [4,5,6]. De novo mHSPC is characterized by an aggressive course with shorter time of onset of castration resistance and worse overall survival (OS) compared to metachronous mHSPC [7]. Since 2015, the prognosis of mHSPC has slightly improved thanks to the introduction of new hormonal agents (NHA) and chemotherapy combined with androgen deprivation therapy (ADT) from the first-line setting [8,9,10,11,12,13,14]. Nonetheless, the current therapeutic decision-making in mHSPC, unlike in metastatic castration-resistant PC (mCRPC), is still based on clinical features (e.g., high-volume vs. low-volume disease, visceral vs. bone-only metastasis) considering that clinical trials evaluating molecular biomarker-guided treatment of mHSPC are still ongoing.

The aim of this review was to describe the current therapeutic opportunities in de novo mHSPC, focusing on potential molecular biomarkers identified in the main clinical trials that have changed the standard of care (SOC), the genomic features of de novo mHSPC, and the principal ongoing trials that are investigating new therapeutic approaches and the efficacy of a biomarker-guided treatment in this setting.

2. Current therapeutic opportunities in de novo mHSPC

2.1. Doublet therapy

2.1.1. Docetaxel plus ADT

The first study that redefined the treatment paradigm in mHSPC has been the CHAARTED (ChemoHormonal Therapy versus Androgen Ablation Randomized Trial for Extensive Disease in Prostate Cancer) trial [8]. This randomized phase III trial study enrolled 790 men affected by mHSPC (575 of them with de novo disease) with the aim to verify the superiority of upfront docetaxel 75 mg/mq given every 3 weeks for six cycles in association with ADT over ADT alone. After a median follow-up of 53.7 months, an absolute benefit in terms of mOS of 16.8 months was observed in the combination treatment arm compared to ADT alone (mOS: 51.2 vs. 34.4 months; HR: 0.63; 95%CI: 0.50-0.79; p<0.001) in patients with high-volume disease (defined by at least 4 bone metastatic lesions with at least one beyond the vertebral bodies and pelvis and/or by the presence of visceral metastases), while no benefit was observed in patients with low-volume disease [15]. A transcriptional profiling of primary PC samples obtained from 160 patients enrolled in this trial (of which 88% with de novo mHSPC and 78% with high-volume disease) was performed by Hamid and colleagues [16] using the PAM50 classifier (luminal A, luminal B and basal subgroups), the Decipher genomic classifier, and androgen receptor activity (AR-A, classified as average vs. lower) [17,18,19]. The analysis revealed a predominance of luminal B (50%) and basal (48%) subtypes, lower AR-A and high Decipher risk disease. Luminal B subtype benefited significantly from the addiction of docetaxel to ADT in terms of OS, in contrast to basal subtype which showed no OS benefit even in case of high-volume disease. In multivariate analysis, higher Decipher risk and lower AR-A significantly associated with poorer OS. Additionally, the combination therapy conferred greater improvements in OS in presence of higher Decipher risk. This study proposed both prognostic and predictive roles for transcriptional subtyping in mHSPC.

2.1.2. Abiraterone plus ADT

The double-blind phase III trial LATITUDE [10] was the first study to demonstrate the benefit of an upfront combination therapy with an NHA. A total of 1199 patients affected by de novo high-risk mHSPC, defined by at least two of the three risk factors (Gleason score ≥ 8, at least three bone lesions, and the presence of visceral metastasis), were 1:1 randomised to receive abiraterone acetate plus prednisone (or prednisolone) plus ADT versus placebo plus ADT. Considering the notable benefit in terms of radiological progression free survival (rPFS) and OS observed in the experimental group at an interim analysis, the trial was subsequently unblinded and crossover was allowed. At the time of final OS analysis (median follow-up 51.8 months), 72 patients had crossed over to abiraterone acetate from the placebo group; the mOS was 53.3 months (95%CI: 48.2 months to not reached (NR)) in the experimental arm vs. 36.5 months (95%CI: 33.5-40.0 months) in the ADT alone arm (HR: 0.66, p<0.0001) [20]. No analysis on predictive biomarkers of response to abiraterone acetate was reported. An interesting multivariable model using data from the LATITUDE trial identified 11 prognostic variables commonly assessed in clinical practice (performance status, number of skeletal metastases, Gleason score, presence of liver metastasis, worst pain score, albumin, LDH level, PSA level, haemoglobin level, and treatment regimen) that accurately predict prognosis and improve risk stratification in de novo mHSPC [21].

2.1.3. Enzalutamide plus ADT

The role of the NHA enzalutamide associated with ADT as upfront therapy in mHSPC has been investigated in two phase III clinical trials. In the double-blind ARCHES trial [13], a total of 1150 patients with mHSPC were 1:1 randomly assigned to receive enzalutamide plus ADT or placebo plus ADT. Previous treatment with docetaxel was allowed. Enzalutamide significantly reduced the risk of radiographic disease progression or death compared with ADT alone by 61% (HR: 0.39; 95%CI: 0.30-0.50; p<0.001) irrespectively of prior local and systemic treatment, disease volume, and risk [22]. A post hoc analysis demonstrated the clinical benefit of enzalutamide in both cases of de novo mHSPC and metachronous mHSPC [23]. After unblinding, 180 progression-free patients randomly assigned to placebo plus ADT crossed over to enzalutamide plus ADT. The final prespecified analysis of OS (median follow-up 44.6 months) showed that enzalutamide reduced risk of death by 34% versus ADT alone (median NR in either group; HR: 0.66; 95%CI: 0.53-0.81; p<0.001) [24].

In the ENZAMET study [12], a total of 1125 patients affected by mHSPC were 1:1 randomly assigned to receive enzalutamide or a non-nonsteroidal first-generation antiandrogen (bicalutamide, flutamide, or nilutamide) in association with ADT. The 52% of the patients had high-volume disease. Prior course of six cycles of docetaxel was given to 65% of patients in the enzalutamide group and to 76% of patients in the control group. Besides, concurrent upfront docetaxel was permitted after a protocol amendment early in accrual. At the planned primary OS analysis (median follow-up 68 months), the mOS was NR in both groups (HR: 0.70; 95%CI: 0.58-0.84; p<0.0001), with 5-year OS of 57% in the control group and 67% in the enzalutamide group. Enzalutamide benefit in terms of OS was consistent across predefined prognostic subgroups (de novo vs. metachronous mHSPC, high-volume vs. low-volume disease) and in those who received concurrent docetaxel [25]. Unfortunately, no analysis on predictive biomarkers of response to enzalutamide was described.

2.1.3. Apalutamide plus ADT

The efficacy of the NHA apalutamide plus ADT compared to ADT plus placebo was evaluated in the double-blind phase III trial TITAN [14]. Eligible patients were required to have mHSPC with at least one lesion detectable on bone scanning; previous docetaxel therapy was allowed. Among 1052 enrolled patients, 10.7% had received previous docetaxel therapy and 62.7% had high-volume disease; more than 80% of patients had de novo disease. The 40% of the patients in the placebo group crossed over to apalutamide after the initial unblinding at 22.7 months of follow-up. At a median follow-up of 44 months, apalutamide plus ADT significantly reduced the risk of death by 35% compared to ADT alone (mOS: NR vs. 52.2 months; HR: 0.65; 95%CI: 0.53-0.79; p<0.0001) and by 48% after adjustment for crossover. Subgroup analysis suggested benefit for apalutamide in almost all subgroups including both low and high-volume disease; a trend towards favouring placebo in patients who had received prior docetaxel was registered, although these patients represented only 10% of the trial population and a post-hoc interaction test showed not interaction between the efficacy of apalutamide and prior docetaxel [26]. In a post hoc analysis, similarly to what performed by Hamid et al. [16], a transcriptional profiling of primary PC samples from 222 patients enrolled in TITAN revealed that most patients had a high Decipher risk disease. In the placebo group, patients with high Decipher risk had poorer prognosis than those with Decipher risk low to average, while no prognostic difference among these two different classes of risk was observed in the apalutamide group. Both basal and AR-A low subtypes showed significant benefit from apalutamide, suggesting that apalutamide is beneficial especially for the highest risk molecular subtypes [27]. A more sustained benefit with the addition of apalutamide to ADT in patients with a high Decipher genomic classifier score compared with patients with a low genomic classifier score was also confirmed in a cohort of 233 patients from the SPARTAN trial [28]. In an exploratory analysis investigating relationships between biomarkers and OS in TITAN, the presence of circulating tumoral DNA (ctDNA) or any androgen receptor (AR) genomic aberrations at baseline and any AR genomic aberrations or PI3K pathway activation at end of study treatment were significantly associated with poor OS in multivariate analyses from both treatment groups [29].

2.2. Triplet therapy

More recently, the efficacy of further treatment intensification with triplet therapy, consisting in the association of ADT with both docetaxel and NHA, has been investigated by the phase III trials ARASENS and PEACE-1. ARASENS [30] enrolled 1306 patients affected by mHSPC eligible for ADT and chemotherapy with docetaxel to receive either darolutamide or placebo in addition to docetaxel for six cycles and ADT. Most patients (86.1%) had de novo mHSPC. The primary analysis showed a 32.5% (HR: 0.68; 95%CI: 0.57-0.80; p<0.001) lower risk of death in the darolutamide group than in the placebo one: with a median follow-up of 43.7 months in the darolutamide arm and 42.4 months in the placebo arm, mOS resulted NR in the experimental group vs. 48.9 month in the control group. According to safety analyses, adverse events (AEs) of any grade were similar in both groups: the most common grade 3 or 4 AE was neutropenia associated with docetaxel. Post hoc analyses showed significant OS benefit in favour of the addition of darolutamide in all patients, with more consistent outcomes in high-volume (mOS: NR vs. 42.4 months; HR: 0.69; 95%CI: 0.57-0.82), high-risk (mOS: NR vs. 43.2 months; HR: 0.71; 95%CI: 0.58-0.86) and low-risk (mOS: NR vs. NR; HR: 0.62; 95%CI: 0.42-0.90) disease subgroups [31]. However, most of the patients included in the ARASENS trial had high-volume (77%) and/or high-risk (70%) disease: low-volume population was not well represented (only 23%). Thus, it is not possible to draw definitive conclusions for patients with low-volume disease.

PEACE-1 [32] was a 2 × 2 factorial design trial which enrolled 1173 patients with de novo mHSPC. Eligible participants were therefore randomly assigned in a 1:1:1:1 to receive the SOC (ADT alone or with docetaxel for six cycles; the 2017 amendment made the association of both mandatory), SOC plus external beam radiotherapy (EBRT) to the primary tumour, SOC plus abiraterone in association with prednisone, or SOC plus abiraterone and EBRT to the primary tumour. In order to assess the efficacy of abiraterone in addition to SOC, on the basis of the assumption of the absence of significant interactions between abiraterone and EBRT to the primary tumour, they conducted a 2X2 factorial analysis. They pooled the groups 2 by 2, distinguishing those who received abiraterone with or without EBRT to the primary tumour into one and comparing them to those who did not receive it (SOC with or without EBRT to the primary tumour). At a median follow-up of 3.5 years, the addition of abiraterone significantly improved median rPFS (4.46 vs. 2.22 years; HR: 0.54; 95%CI: 0.41-0.71) with a reduction of the relative risk of radiographic progression by 46%. With a median follow-up of 4.4 years, also a significant benefit in terms of mOS was reported in patients receiving abiraterone (5.72 vs. 4.72 years; HR: 0.82; 95%CI: 0.69-0.98; p=0.03), with a risk of death from any cause 18% lower than in those who didn’t receive it. The effect of abiraterone was particularly marked in patients with high-volume disease (median rPFS: 4.46 vs. 2.03 years, HR: 0.50; mOS: NR vs. 4.43 years, HR: 0.75). From the safety point of view, abiraterone did not determine a significant increase in neutropenia, febrile neutropenia, fatigue, or neuropathy rates compared with ADT plus docetaxel alone; the only exceptions were hypertension, hypokalaemia, and higher levels of aminotransferases, which were more frequently reported in the group treated with abiraterone.

Both ARASENS and PEACE-1 showed that upfront treatment intensification with the combination of ADT, docetaxel, and NHA in de novo mHSPC could become a new SOC since it improved survival outcomes with an acceptable safety profile, especially in patients with high-volume disease. Up to now, no predictive biomarker of response to triplet therapy has been reported.

2.3. Oligometastatic prostate cancer

Oligometastatic PC (omPC) encompasses a heterogeneous group of tumours characterized by a low metastatic burden [33]. While some works define omPC on the basis of the number of metastases, ranging from 3 to 5 lesions, other authors adopt the criteria of low-volume disease according to CHAARTED trial [8] or low-risk disease according to LATITUDE trial [10] for the definition of omPC either de novo or recurrent [34]. Considering that de novo omPC has generally an indolent behaviour, with lymph node metastases only or minimal bone involvement, and it is associated with a better prognosis compared to patients with more than 5 lesions [35], a benefit from different treatment options may be observed. In fact, post hoc analysis of CHAARTED [8] and GETUG-AFU15 trials [36] showed that patients with low-volume disease had a much longer OS, without evidence that docetaxel improved OS, irrespective of whether patients received ADT plus docetaxel for de novo mHSPC or after prior local treatment [37]. By the other hand, in a post hoc analysis of the STAMPEDE trial arm G [11], the addition of abiraterone to ADT improved OS also in low-volume de novo mHSPC (HR: 0.60, 95%CI: 0.39-0.92) [38]. Similarly, upfront enzalutamide or apalutamide conferred a disease burden-independent advantage over ADT alone in the phase III pivotal studies [12,13,14].

Different therapeutic approaches in de novo omPC include locoregional treatments, mainly radiation therapy. In the HORRAD trial [39] 432 patients with primary bone mHSPC were randomised to receive ADT alone or ADT plus EBRT to the primary tumour; the subgroup analysis demonstrated a trend towards an OS benefit only in patients with less than 5 bone lesions (HR: 0.68, 95%CI: 0.42-1.10). These promising results were further investigated in the STAMPEDE trial arm H [40]: EBRT to the primary tumour significantly improved OS in patients with low metastatic burden according to CHAARTED criteria (HR: 0.68, 95%CI: 0.52-0.90; p = 0.007), reporting an increase of the 3-year survival rate from 73% to 81% with EBRT. In a recent phase II trial including 200 patients with de novo omPC (defined as 5 or fewer bone or extrapelvic lymph node metastases and no visceral metastases) randomised to receive either ADT or ADT plus radical local treatment of the primary tumour, both rPFS and OS were significantly improved in the arm with radical local treatment of the primary tumour [41]. However, opposite results have been recently presented at the last ASCO genitourinary symposium from the PEACE-1 trial [30]: in men with de novo low-volume mHSPC (at most 3 bone metastases with or without nodal involvement) combining prostate EBRT to systemic treatment did not improve OS [42]. The differences emerged in these trials are probably due to the different definition of low-volume diseases as well as the different systemic treatments administered to the patients. Nevertheless, EBRT to the primary tumour combined with the systemic treatment is recommended for patients with low-volume mHSPC according to ESMO and NCCN guidelines [43,44].

In addition to EBRT to the primary tumour, metastasis-directed therapy (MDT) is a debating issue. MDT is generally used to treat bone metastases or pathological lymph nodes. The only two prospective trials of stereotactic ablative radiotherapy (SABRT) versus observation, STOMP and ORIOLE, were focused only on metachronous omPC, demonstrating that MDT prolongs androgen deprivation-free survival and PFS compared to observation [45,46]. Although in de novo omPC there is no randomised trial evidence suggesting a benefit from MDT of all documented lesions, there is a strong consensus for a combined approach (ADT plus additional systemic therapy, local radiotherapy, and MDT) [43]. Available evidence derives from several case series in which a combined approach was investigated with encouraging results [48,49,50,51]. Many trials are ongoing to define whether the combination of ADT plus SABRT in de novo omPC improves outcomes compared with systemic treatment alone (NCT03298087, NCT05707468, NCT04983095, NCT04115007, NCT05223803, NCT04619069, NCT03784755, NCT05212857, NCT05209243).

The addition of radiation therapy to systemic treatment has a potential biological rationale: radiotherapy induces cell death, and the dying cells release “danger signals” that in turn might make cancer cells outside the radiation field more susceptible to an immune-mediated cytotoxic environment (the so-called abscopal effect) [52]. Moreover, radiation therapy might prevent metastasis-to-metastasis spread. Characterizing multiple metastases arising from PC in ten patients under ADT with whole-genome sequencing, Goundem et al. [53] demonstrated the existence of metastasis-to-metastasis spread, either through de novo monoclonal seeding of daughter metastases or through the transfer of multiple tumour clones between metastatic sites.

Although MDT appears to be effective in omPC, little is known about predictive biomarkers of response to the different treatment options available in this setting [54,55]. The study of predictive biomarkers might be useful to identify which patients could benefit from ADT only or ADT combined with chemotherapy, NHA and/or local treatments. The only data available derive from a pooled analysis of STOMP and ORIOLE trials, where the largest benefit of MDT in metachronous omPC was observed in patients with high-risk mutations defined as pathogenic somatic mutations within ATM, BRCA1/2, Rb1, or TP53, suggesting that a high-risk mutational signature may stratify treatment response after MDT [56].

3. Genomic features of mHSPC

The aim of therapy modulation and personalization in de novo mHSPC may be reached by the study of biology and biomarkers. However, the mutation profile of mHSPC is poorly characterized since sequencing efforts have focused on either localised PC or mCRPC. Progression from localized PC to mCRPC is characterised by accumulation of deleterious genomic alterations in the latter disease state. In detail, the most frequent altered genes in mCRPC are tumour suppressor genes (RB1, TP53 and PTEN) and genes involved in androgen receptor (AR) pathway, chromatin remodelling (KMT2C and KMT2D), PI3K signalling (AKT1 and PIK3CA), and DNA damage repair (DDR) (BRCA2, BRCA1, ATM, and FANCA) [57,58,59]. Different data seems to indicate that the mutation profile of mHSPC lies between localised PC and mCRPC, suggesting that the enrichment of deleterious alterations over time confers survival advantage to cancer cells inducing treatment resistance [60,61]. A systematic metanalysis [54] including 1682 mHSPC patients of whom 1248 (74%) with de novo disease from 11 studies pointed out that the most commonly altered genes, due to mutations or copy number alterations, were TP53 (32%) and PTEN (20%), followed by genes involved in DDR (18%) with BRCA2 as the most frequently mutated gene (7%); alterations in cell cycle signalling were reported in 7-13% of the cases. Tumours from patients with de novo mHSPC were enriched for mutations in TP53 and CDK12 compared with metachronous mHSPC, while cell cycle signalling, Wnt pathway, PTEN and SPOP alterations were more frequent in metachronous mHSPC. In high-volume disease according to CHAARTED criteria [8], TP53, BRCA2, PIK3CA, RB1, and APC were more frequently altered compared to low-volume disease. However, DNA source and definitions for genes alterations differed significantly among studies, including somatic alteration from formalin-fixed paraffin-embedded (FFPE) material as well as ctDNA. Among the studies included in the aforementioned metanalysis, a noteworthy observation derives from the targeted next-generation sequencing (tNGS) performed on 185 tumour samples obtained almost entirely from de novo mHSPC patients enrolled in the STAMPEDE trial: PI3K pathway aberration was observed in 43% of the cases, due to PTEN copy-number loss (34%) and/or inactivating mutations in PIK3 or AKT (18%) [62].

3.1. The role of liquid biopsy

Most patients with de novo mHSPC will not undergo primary surgery; the histologic diagnosis is typically performed on prostatic biopsy. Therefore, liquid biopsy could add clinically relevant information in this setting. In a single-centre prospective cohort, Vanderkerhove et al. [63] detected a median plasma ctDNA fraction of 11% (range 2.0-84%) among 26 out of 35 (74%) untreated patients with de novo mHSPC; for the remaining 9 patients, ctDNA was not detectable. Higher ctDNA levels were identified in presence of visceral metastasis. The somatic analysis from ctDNA and tumour tissue revealed a mutational landscape similar to mCRPC, although without AR gene alterations: TP53 and DDR genes mutations were identified in 47% and 21% of the cases, respectively. The rate of concordance for mutation detection between tumour tissue and ctDNA was 80%, suggesting that de novo mHSPC is a highly clonal disease at the diagnosis. By the other hand, in a cohort of 82 Chinese patients with de novo mHSPC, only 50% of patients had a ctDNA fraction >2% and the percentage of ctDNA positive patients was even lower (37%) in a cohort of 73 untreated mHSPC including both de novo and metachronous disease [64,65]. There are still challenges to overcome prior to introduce liquid biopsies in routine clinical practice, such as preanalytical aspects and low-circulating tumour content, considering that common PC copy number alterations such as PTEN or CDH1 deletions are undetectable in presence of low ctDNA fraction [66]. Consequently, in patients with a low ctDNA fraction, tissue biopsy profiling remains more informative.

3.2. Prognostic information

From a prognostic point of view, data regarding the association between genetic alterations, time to castration resistance and OS in de novo mHSPC are partial, because they have been obtained from cohorts including both synchronous and metachronous metastatic disease. Among 424 cases of mHSPC including 275 patients with de novo disease, Stopsack et al. [67] reported a rate of progression to castration resistance 1.6 to 5-fold higher in presence of alterations in AR, TP53, cell cycle, and MYC pathways and approximately 1.5-fold lower with SPOP and Wnt pathway alterations; similarly, OS rate was 2 to 4-fold higher in presence of AR or cell cycle alterations, and 2 to 3-fold lower if SPOP or Wnt pathway was altered. Sequencing of FFPE tissue from biopsies of 43 patients affected by mHSPC of whom 30 with de novo disease revealed an incrementally poorer OS with cumulative mutations or alterations in the tumour suppressor genes TP53, PTEN, and RB1 [61]; the negative prognostic value of alterations in TP53, PTEN, and RB1 has been observed also in a cohort of 97 patients with mHSPC treated with first-line ADT plus docetaxel or abiraterone acetate, outperforming clinical criteria to predict early disease progression [68]. An association between shorter OS and alteration in TP53, ATM and DDR genes detected on plasma ctDNA was observed also among 53 patients with de novo or metachronous untreated mHSPC [65]. Finally, tNGS across 113 genes performed on 202 primary tumours samples obtained from patients with synchronous or metachronous mHSPC revealed a significantly shorter OS in presence of mutations or deep deletions of RB1 [69].

The association between SPOP mutations and better prognosis has been also detected in a cohort of 121 men with de novo mHSPC treated with ADT: both median PFS and OS were significantly improved in the subset of 25 patients with SPOP mutated cancers (mPFS: 35 vs. 13 months, p = 0.016; mOS: 97 vs. 69 months, p = 0.027) [70]. SPOP protein is involved in ubiquitination and consequent proteasomal degradation of target proteins; in PC, SPOP seems to act as a tumour suppressor by targeting several proteins, including AR, SRC3 and BRD4 [71]. The hypothesis that SPOP mutated PC are primarily driven by AR signalling has been tested in a real-world setting: in a cohort of patients with de novo mHSPC undergoing ADT plus NHA, the presence of SPOP mutation compared with wild-type was associated with longer time to castration resistance and OS, while SPOP mutational status was not associated with time to castration resistance nor OS in a cohort treated with ADT plus docetaxel [72]. SPOP mutation may therefore be used as a predictive biomarker to guide treatment choice for patients with de novo mHSPC.

4. Ongoing phase III clinical trials testing new therapeutic approaches in mHSPC

Apart from trials focusing on NHA, ADT and chemotherapy with different schedules in mHSPC (ARANOTE NCT04736199, ARASAFE NCT05676203, LIBERTAS NCT05884398, NCT05956639), several other ongoing phase III clinical trials are investigating the role of new therapeutic approaches (immunotherapy, radiopharmaceuticals, and molecular target agents) in this setting (Table 1, Figure 1).

4.1. Immunotherapy

Although the expression of programmed death ligands 1 and 2 (PD-L1 and PD-L2) on PC cells is highly variable, therapy with enzalutamide can upregulate PD-L1 expression in the tumour microenvironment; this can represent a mechanism of resistance by inducing immune evasion [73]. In the phase Ib Keynote-028 and phase II Keynote-199 trials, mCRPC enzalutamide-refractory patients and previous untreated patients received the combination of pembrolizumab and enzalutamide, reaching potentially enhanced and durable response rates [74,75]. Based on these premises, the ongoing randomized, double-blind, placebo-controlled, phase III KEYNOTE-991 (NCT04191096) [76] is investigating if this combination therapy in NHA-naive participants with mHSPC is superior to enzalutamide plus placebo. Stratification by prior docetaxel therapy and the presence of high-volume disease is planned. Pembrolizumab 200 mg every three weeks will be administered for up to 35 cycles, loss of clinical benefit or intolerable AEs. The two co-primary endpoints are OS and rPFS. Archival or newly obtained tumour tissue and blood for genetic, RNA, serum, and plasma biomarkers and ctDNA analyses will be collected from all participants to support exploratory analyses of novel biomarkers. PROSTRATEGY (NCT03879122) is another phase III clinical trial that is investigating the role of immunotherapy in high-volume mHSPC [77]. This trial will randomise approximately 135 patients in three arms: ADT + docetaxel for 6 cycles (control arm, ARM 1); ADT + docetaxel for 6 cycles and then nivolumab 3 mg/kg every two weeks for 12 months (ARM 2); ADT + 2 cycles of ipilimumab 3 mg/kg every 3 weeks, followed by 3 cycles of docetaxel, 2 cycles of ipilimumab, 3 cycles of docetaxel, and then nivolumab 3 mg/kg every two weeks for 12 months (ARM 3). The primary endpoint is OS.

4.2. Radiopharmaceuticals

Lutetium-177(177Lu)-PSMA-617 is a beta emitter radioisotopic agent approved by FDA in 2022 for the treatment of mCRPC in patients who had progressed to an NHA and a taxane-based chemotherapy, and whose metastatic lesions express the prostatic-specific membrane antigen (PSMA) as documented by PSMA imaging [78]. Radiopharmaceuticals release alpha or beta radiations to cancer cells thought radioisotopes; radiations activate apoptosis by single- and double-strand DNA breaks [79]. PSMAddition (NCT04720157) [80] is a phase III, randomized, open-label, international, prospective clinical trial that aims to evaluate the efficacy and safety of 177Lu-PSMA-617 in combination with SOC (ADT plus NHA), versus SOC alone, in mHSPC. About 1126 patients will be randomized 1:1 to receive the SOC, with or without 177Lu-PSMA-617 administered once every 6 weeks for six cycles. Exclusion criteria is a rapidly progressing tumour that requires chemotherapy. The primary endpoint is rPFS. Stratification according to age (≥70 years/<70 years), high-volume vs. low-volume disease and previous/planned prostatectomy or radiation treatment of the primary prostate tumour is planned.

4.3. Molecular target agents

The role of molecular target agents has been largely investigated in the mCRPC setting in combination with ADT. The increase in knowledge of the mutational profile in mHSPC is leading to test targeted treatments, such as cyclin-dependent kinase 4 and 6 (CDK4/6) inhibitors, poly(ADP-ribose) polymerase inhibitors (PARPi), and AKT-inhibitors (AKTi), also in this setting [81].

4.3.1. CDK4/6 inhibitors

During G1-S checkpoint, CDK4/6 activation by AR axis contributes to cancer cell proliferation; among mechanisms of resistance to NHA, upregulation of cyclin D1 (whose association with CDK4/6 is crucial for transition from G1 to S phase) has been described [82]. CYCLONE-03 (NCT05288166) [83] is a placebo-controlled phase III study that will randomise about 900 patients affected by high-risk NHA-naïve mHSPC (defined by at least 4 bone metastasis and/or visceral disease) to receive either abemaciclib (a selective CDK4/6 inhibitor) or placebo, plus abiraterone and prednisone. Visceral metastases and de novo mHSPC are stratification factors. The primary endpoint is rPFS.

4.3.2. PARP inhibitors

Preclinical and clinical data have showed that co-inhibition of AR axis and PARP induces a combined anti-tumour effect: PARP is involved in positive co-regulation of AR signalling, so PARP/AR signalling co-inhibition leads to enhanced AR target gene suppression; moreover, treatment with NHAs inhibits the transcription of some DDR genes, inducing synthetic lethality by cancer cells’ inability to repair DNA even in patients without any DDR alterations [84]. The combination of the PARPi olaparib with abiraterone is FDA and EMA approved as first-line treatment of mCRPC, if chemotherapy is not clinically indicated, according to the results of PROPEL [85]. The combination of the PARPi talazoparib and enzalutamide has been recently FDA approved as first-line treatment of patients with homologous recombination repair (HRR) gene-mutated mCRPC according to TALAPRO-2 [86]. TALAPRO-3 (NCT04821622) [87] is a randomized double-blind trial that has recruited 599 men with mHSPC and HRR-related genes alterations (ATM, ART, BRCA1, BRCA2, CDK12, CHEK2, FANCA, MLH1, MRE11A, NBN, PALB2, RAD51C) to receive enzalutamide in association with placebo or talazoparib. The primary endpoint is rPFS. Patients will be stratified according to non-BRCA vs. BRCA alteration, low-volume vs. high volume disease, de novo vs. metachronous mHSPC. Similarly, the randomized, placebo controlled, double-blind trial AMPLITUDE (NCT04497844) [88] will enrol approximately 692 patients with mHSPC and HRR alterations to receive the PARPi niraparib or placebo in combination with abiraterone. The primary endpoint is rPFS. Patients will be stratified according to disease volume, previous docetaxel-based chemotherapy, and the type of HRR-related gene defect.

4.3.3. AKT inhibitors

Inactivation of the tumour suppressor gene PTEN by deletion or mutation is frequent in PC, especially in late-stage tumours. Loss of PTEN function leads to PI3K/AKT signalling pathway activation and suppression of AR transcriptional output. AKTi activate AR signalling, suggesting potential efficacy of the inhibition of both PI3K and AR signalling pathways [89]. Evidence supporting this association comes from the phase III trial IPATential150 [90] that demonstrated that the AKTi ipatasertib, in association with abiraterone, improve rPFS in patients with mCRPC and PTEN-loss. CAPItello-281 (NCT04493853) [91], a randomized double-blind trial, will test the AKTi capivasertib. Approximately 1000 patients with mHSPC PTEN-deficient, demonstrated on tissue immunohistochemistry, will be randomized 1:1 to receive capivasertib or placebo in association with abiraterone. The primary endpoint is rPFS.

5. Conclusions

The road towards a personalized treatment for de novo mHSPC is still long considering that the randomized clinical trials, which have furnished the basis of the current therapeutic options, stratified patients according to clinical criteria that not necessarily reflect the biological rationale of the chosen therapy. Transcriptomic profiling of mHSPC has revealed a predominance of aggressive and poor prognosis subtypes, but its role as a predictive biomarker requires further validation. Even though many of the genomic alterations detected in mHSPC are considered predictive in mCRPC, it remains to ascertain how these alterations can be exploited in the mHSPC field. In this sense, the ProBio (NCT03903835) trial, that is randomizing both mHSPC and mCRPC to receive SOC following national guidelines (control arm) or treatments based on a biomarker signature inferred from diagnostic tissue or liquid biopsy profiling (experimental arm), will probably furnish prospective evaluation of biomarker-driven treatments.