In recent decades, a large number of novel drugs with different mechanisms of action have been introduced in the therapeutic management against prostate cancer in all stages, focusing mainly on metastatic and castration-resistant disease, with and without metastasis, where they have shown to delay progression and prolong survival. Clinical options include cytostatic agents (docetaxel plus prednisone, cabacitaxel), second-generation antiandrogens, (abiraterone, enzalutamide, apalutamide, and darolutamide) bone-targeted therapies (Radium 223), immunotherapy (sipuleucel-T), Akt inhibitors, and radioisotopes (Lutetium-177–PSMA-617). However, there is a lack of high-quality randomized trials exploring different treatment sequences.
Therefore, there is an urgent need for more effective treatments for patients with mCRPC, especially those who have progressed after novel hormonal therapies and taxane chemotherapies.
Synthetic lethality is a genetic concept in which the functional loss of two genes results in cell death, while the functional loss of only one of them is compatible with cell viability (Figure 1).
BRCA1 and BRCA2 are tumor suppressor genes involved in transcriptional regulation and repair of double-strand DNA damage breaks (DSBs) in the DNA molecule, playing a key role on the HR path. Cells with loss-of-function in these genes are unable to repair errors in DNA, depending on the ability of PARP to activate alternative routes. Since they depend on PARPs to maintain genome integrity, these cells are highly vulnerable to PARPi.
2. Rationale of Use of PARP Inhibitors in PCa
2.1. DNA Repair Pathway
The integrity of DNA is continuously challenged by a variety of agents and processes that can alter, directly or indirectly, the sequence of this molecule. Damage that arises in DNA, its repair, or its absence are key factors in the appearance of mutations that initiate and promote tumorigenesis
[9].
Given the potentially devastating effect of these types of mutations, the healthy cells evolved to defend themselves from the deleterious effects that DNA damage causes through several molecular pathways, which, as a whole, are called DNA Damage Response (DDR). These pathways have the ability to identify the damage, promote the arrest of the cell cycle, and, ultimately, proceed with its repair, contributing to the maintenance of genome integrity.
DDR can be divided into distinct but functionally interconnected means that process the repair of existing damage: first, the repair of double-strand DNA damage breaks (DSB), which includes repair by Homologous Recombination (HR) and nonhomologous recombination; second, the repair of single-strand breaks (SSB), of which the excision repair base (ERB) is a part. Several key proteins contribute to the correct function of these processes. The BRCA1, BRCA2, PALB2, ATM, CHEK1, CHEK2, and RAD51 play a role in the HR process. However, Poly enzymes (ADP-Ribose) Polymerase 1 and 2 (PARP1 and PARP2) are the pillars of the ERB process
[10].
Those proteins also play a key role in the repair of DSB by facilitating the activation of the repair by HR and contributing to the inhibition of less conservative repair pathways, such as the nonhomologous route end joining (NHEJ). Thus, the absence of PARPs contributes to the dysfunctional HR process, making nonconservative DNA repair processes the dominant pathways
[11].
2.2. Role of DNA Damage Repair Genes in Prostate Cancer
2.2.1. DDR Mutations in Prostate Cancer
Classical studies show that germline mutations in HR DNA repair genes have been observed in approximately 10–15% of patients with metastatic prostate cancer, while somatic mutations occur in 20–25% of those patients
[12][13].
Of these mutations, BRCA2 mutations are the most frequently found (12–18%), while ATM (3–6%), CHEK2 (2–5%), and BRCA1 (<2%) are other commonly altered genes involved in HRR
[14][15].
Since DDR pathway alterations were seen at similar rates between localized and metastatic PCa, it has been speculated that PARPi may also have a therapeutic effect in localized PCa.
To support this hypothesis, a 2019 study by Kim et al.
[16] analyzed the DDR pathway alterations in localized PCa using The Cancer Genome Atlas (TCGA) public database. Their results highlighted that the DDR alteration rate was surprisingly higher than suggested by previous studies
[17][18] and was associated with shorter OS in men with postoperative HR features.
A systematic review of the p`revalence of DNA-damage-response gene mutations in prostate cancer that summarizes the prevalence of DDR mutation carriers in the unselected (general) PCa and familiar PCa populations confirms these findings
[19].
BRCA1/2 genes are located at chromosomes 17q21 and 13q12, respectively. They are large genes consisting of 100 and 70 kb, respectively. They have an autosomal dominant inheritance pattern with incomplete penetrance.
2.2.2. Germline Mutations
Multiple studies have reported an association between frequent germline deleterious mutations in DDR genes and advanced PCa. Specifically, germline BRCA1/2 mutations are associated with increased risk and more aggressive PCa, higher risk of nodal involvement, and distant metastasis at diagnosis
[20].
Germline BRCA2 mutations increase the risk of developing PCa eightfold at the age of 65 years
[21]. In the localized disease, germline BRCA1/2 mutations are related with progression in patients undergoing active surveillance and a high rate of recurrence in patients treated with curative intention
[22][23]. Similarly, in the metastatic setting, it is associated with a more aggressive evolution
[24].
The prevalence of germline mutations varies among countries and ethnic groups.
The International Stand Up to Cancer/Prostate Cancer Foundation team (SU2C-PCF) studied 150 metastatic PCa identifying 8% of germline and 23% of somatic DDR mutations
[12]. BRCA2 was the most frequent mutation (13%), followed by ATM (7.3%), MSH2 (2%), and BRCA1, FANCA, MLH1, RAD51B, and RAD51C.
Pritchard et al. studied germline mutations in 692 men with mCRPC with no family history, identifying 84 deleterious mutations in the 20 DNA repair genes studied in 82 men (11.8%), being BRCA2 the most frequent one (5.3%)
[13].
Nicolosi et al. analyzed 3607 men with PCa
[25], finding germline mutations in 620 of them (17.2%), of which 30.7% were BRCA1/2 mutations; other mutations included ATM, PALB2, CHEK2, and mismatch repair genes PMS2 and MLH1/2/6.
2.2.3. Somatic Mutations
Somatic mutations also stimulate carcinogenesis
[26]. As explained before, Robinson et al. found that 23% of mCRPC had somatic mutations in DNA repair pathway genes, with most mutations in BRCA2 and ATM
[10]. Different studies have found 12% and 8% of PCa patients carrying a BRCA1/2 or ATM mutation, respectively, and more frequently in mCRPC
[27]. Abida et al. observed somatic BRCA2 mutations in tumors before they progressed to metastatic disease. Somatic BRCA2 mutations have been suggested to arise early in tumors from patients who eventually develop metastatic disease, while ATM alterations seem to enrich in CRPC
[15].
2.3. Mechanism of Action of PARP Inhibitors
The original rationale for using PARPi as a cancer treatment is that PARPi can sensitize tumor cells to therapies that cause DNA damage, such as chemo or radiotherapy. The inhibition of PARP-mediated repair of DNA damage produced by chemotherapy or radiotherapy, may result in an increased of therapeutic potency
[28].
Almost two decades ago, two groups described the important concept of Synthetic Lethal (SL) interaction between PARP inhibition and BRCA1 or BRCA2 mutation, which represented a new therapeutic option for BRCA-mutant tumors
[29][30].
SL means that a defect in either one of two genes has little effect on the organism, but a combination of defects in both genes results in cell death.
Carriers of deleterious heterozygous germline mutations in the BRCA1 and BRCA2 genes have high risk of different types of cancer, such as PCa
[31]. BRCA1 and BRCA2 are tumor-suppressor genes involved in transcriptional regulation and, as stated before, are critical to the repair of DSBs in the DNA molecule, playing a key role in the HR pathway
[30]. Cells with functional loss in these genes are unable to repair errors in DNA, depending on PARPs ability to detect these damages and activate alternative repair pathways.
By depending on PARPs to maintain genome integrity, these cells are highly vulnerable to PARPi. Although the functional loss of only one of the genes is tolerated (BRCA or PARP), the inhibition of PARP in cells with BRCA mutations makes them unable to repair DNA damage, causing the accumulation of errors and, ultimately, leading to cell death (synthetic lethality)
[10].
PARPi antitumor activity is based on the concept of synthetic lethality, in which two separate molecular pathways, which are not lethal when disrupted individually, cause cell death when inhibited simultaneously (Figure 2).
Figure 2. PARPi antitumor activity based on Synthetic Lethality. I: Normal cell. Excision repair base (ERB) and homologous recombination (HR) are functional: repair DNA maintaining cell viability. II, III: Through BRCA mutation or PARPi, one of the repair pathway’s is inhibited. Once the other pathway is functional, the cell maintains its viability. IV: Both DNA repair pathways are inhibited; therefore, errors in the DNA are not repaired, resulting in cell death.
The demonstration that BRCA-mutant tumor cells were as much as 1000 times more sensitive to PARPi than BRCA-wild type cells provided the impetus for PARPi to be tested in clinical trials as single agents. PARPi were the first clinically approved drugs to explore the SL mechanism
[32]. This concept has been reviewed, since it has been shown that some PARPi “trap” PARP1 on DNA, preventing autoPARylation and PARP1 release from the site of damage, thereby interfering with the catalytic cycle of PARP1
[33][34][35].
PARPi differ in their capacity to trap PARP1; talazoparib is nearly 100 times more potent than niraparib, followed by olaparib and rucaparib. However, veliparib has a low ability to trap PARP1. These differences in PARP1 trapping are a predictor of in vitro cytotoxicity in BRCA mutant cells, being talazoparib the one with more cytotoxic capacity
[34] (
Figure 3).
Figure 3. PARPi Trapping Potency.
This different trapping capacity is also important when combined therapies are studied, since a high trapping activity could be accompanied by high toxicity.
Initially, it was postulated that PARPi would only be effective in tumors with mutations in the BRCA1/2 genes. However, in more recent clinical studies, it was found that tumors without BRCA1/2 mutations might respond to this treatment, although its benefits are less significant
[36][37].
The “trapping” effect of PARPi can also have a direct cytotoxic effect: PARPi will trap PARP proteins to the ssDNA damage sites, preventing its release, which is essential for mediator proteins to initiate repair. The trapped PARP proteins will cause replication fork stalling, which will increase replication stress and eventually cause cell death
[38].
Therefore, based on these observations, the number of patients who could theoretically benefit from PARPi has been widely extended. PARPi can be used as monotherapy or in association with other drugs, namely in combination with chemotherapy, immunotherapy, and other targeted therapies that limit DNA damage repair. The combination of drugs has a synergistic and advantageous effect, to the extent that it makes it possible to overcome the resistance to PARPi and increase the effectiveness of these drugs
[39].
Table 1 represents the chemical structures and pharmacological properties of the currently available PARP inhibitors (PARPi).
Table 1. Chemical structures and pharmacological properties of the currently available PARP inhibitors (PARPi).
| |
Structural Formula |
Dose [mg] |
Cmax [ng/mL] |
Tmax [h] |
Half-Life [h] |
PARP1 Trapping Ability |
Direct Off-Targets |
Indirect Off-Targets |
| Olaparib (AZD-2281, MK-7339) |
 |
300/12 h |
7700 |
1.5 |
14.9 |
Moderate |
PARP1, PARP2, PARP3 |
|
| Rucaparib (AG014699) |
 |
600/12 h |
1940 |
1.9 |
25.9 |
Moderate |
ARTD5, ARTD6, PARP1,2,3 TNKS1, TNKS2 |
ALDH2, H6PD, CDK16, PIM3, DYRK1B |
| Niraparib (MK-4827) |
 |
300/24 h |
2232 |
3.0 |
36 |
Moderate–high |
PARP2, PARP3, PARP4, PARP12 |
ALDH2, CIT, DCK, DYRK1A, DYRK1B |
| Talazoparib (BMN-673) |
 |
1/24 h |
16.4 |
1–2 |
90 |
Very high |
PARP2 PARP1 |
|
| Veliparib (ABT-888) |
 |
40/12 h |
410 |
1.0 |
6 |
Very low |
PARP2, PARP3, PARP10 |