Exploiting defects in homologous recombination repair for metastatic, castration-resistant prostate cancer

ABSTRACT

The PROfound trial highlights that there is a benefit in testing genes involved in homologous recombination (HR) and forms the rationale for testing in all patients with metastatic, castration-resistant prostate cancer (mCRPC). This trial also demostrates that olaparib improves progression free survival (PFS), objective response rate (ORR), and time to pain progression in patients who harbor alterations in BRCA1, BRCA2, and ATM. These are groundbreaking findings – this is the first trial that demonstrates the efficacy of olaparib versus standard therapy in a genomically-selected patient population with metastatic prostate cancer. Although this trial does not demonstrate improvements in overall survival (OS), we believe that this may be an underestimation based on trial-design. Future studies of olaparib are likely to yield further promising results.

KEYWORDS:

• Homologous recombination
• DNA repair
• metastatic prostate cancer
• olaparib

Homologous recombination in metastatic prostate cancer

Prostate cancer is one of the most heritable cancers, and 57% of a man’s risk is from genetic factors, as evidenced by a study of monozygotic and dizygotic twins in Nordic countries.¹ In a consortium study of mCRPC, 23% of patients harbor DNA repair pathway aberrations, and 8% have germline changes.² This study also showed that somatic and germline aberrations of BRCA2 exist in 12.7% of cases, of which approximately 90% exhibited biallelic loss. Given the surprisingly high frequency relative to primary prostate cancer, evaluation was expanded to include additional DNA repair genes, which were found in 22.7% of cases. Aberrations were additionally identified in BRCA1, ATM, CDK12, FANCA, RAD51B, and RAD51C. Together, aberrations in BRCA2, BRCA1, and ATM exist in 19.3% of mCRPC patients. Another study showed that 11.8% of patients with metastatic prostate cancer have germline mutations in DNA-repair genes, including BRCA2 (5.3%), ATM (1.6%), CHEK2 (1.9%), BRCA1 (0.9%), RAD51D (0.4%), and PALB2 (0.4%).³ These mutations did not differ based on age of diagnosis or family history of prostate cancer. Importantly, mutations in genes involved in HR repair of DNA have been associated with more aggressive disease with regard to histological grade, risk of recurrence, and prostate cancer-specific death.⁴ Loss-of-function mutations in DNA repair pathway genes predispose to familial prostate cancer with more frequent nodal involvement, metastasis, or T4 stage.⁵

The PROfound trial identified patients with somatic mutations that are involved in the proper functioning of DNA repair machinery. DNA double-strand breaks (DSBs) are deleterious for mammalian cells and can be introduced by ionizing radiation,⁶ various chemical compounds,⁷ and free radicals that are created during metabolism.⁸ DSBs can be repaired by either HR or non-homologous end-joining.⁹

In HR, repair of DSBs occurs by using DNA sequences from sister chromatids as a template for the HR repair machinery. This process is initiated by the formation of the MRE11-RAD50-NBS1 (MRN) complex, which forms a heterodimer with BRCA1-BARD1 and CtIP to convert a DSB into single-stranded DNA (ssDNA). The purpose of this conversion is to allow the RAD51 recombinase to attach to the ssDNA and to form a nucleoprotein filament in a process that is mediated by a protein complex consisting of BRCA2-BARD1-PALB2. RAD51 can then mediate an exchange of DNA strands amongst sister chromatids and allow DNA recombination, thereby allowing the synthesis of missing DNA sequences. The MRN complex also acts as a sensor for the activation of ATM, leading to activation of downstream targets, S-phase checkpoint activation, and survival from genetic insults.^10,11

Rationale for olaparib

BRCA1 and BRCA2 were initially discovered in the context of breast and ovarian cancer susceptibility,^12,13 and can be targeted by olaparib, a PARP inhibitor, by exploiting synthetic lethality.^14–16 Cells that have mutations in BRCA1/2 are deficient in HR and are therefore incapable of repairing DSBs.¹⁷ Additionally, PARP is a DNA repair enzyme, and inhibition of PARP leads to the accumulation of DSBs, which ultimately leads to death of the cancer cell. Olaparib is also effective against pancreatic and prostate cancer.¹⁸ In prostate cancer, BRCA1/2 mutations are more frequently associated with Gleason $\ge$8, T3/T4 stage, nodal involvement, metastases at diagnosis, decreased cause-specific OS (including in localized cases), and decreased metastasis-free survival.¹⁹

Two pioneering trials attempted to further define the role of olaparib therapy in patients with metastatic prostate cancer and alterations in HR-related genes. In the TOPARP-A trial, patients with mCRPC were treated with olaparib, and next-generation sequencing identified deletions and/or mutations in BRCA2 and ATM in those who responded to olaparib.¹⁶ The TOPARP-B trial was a phase II trial that randomized patients with DNA damage response gene aberrations as identified by next-generation sequencing to treatment with olaparib 400 mg or 300 mg twice daily.²⁰ Radiological response was 24.2% v 16.2%, a decrease of PSA by 50% or more was 37.0% v 30.2%, and circulating tumor cell count conversion was 53.6% v 48.1% in the two cohorts, respectively. This was the first prospective trial in a genomically-defined population of metastatic prostate cancer. Unfortunately, neither study contained a standard-therapy arm. The TOPARP-A trial was a single-arm study with olaparib, and both arms of the TOPARP-B trial contained olaparib.

Comparing olaparib to standard therapy

The PROfound study is novel because it exploits synthetic lethality in a group of patients with mCRPC whose disease had progressed during treatment with enzalutamide or abiraterone. Patients who were previously treated with taxane chemotherapy were also included. Eligible patients also had alterations in at least 1 of 15 genes that are central to HR-mediated repair of DNA damage. These patients were randomized (in a 2:1 ratio) to either olaparib or physician’s choice of enzalutamide or abiraterone. This is the first study that both uses a genomically-defined population and employs standard treatment as a control arm in metastatic prostate cancer patients, thus demonstrating that olaparib is a good treatment choice for patients with defects in the HR pathway.

In patients with at least one alteration in BRCA1, BRCA2, or ATM (Cohort A), the primary endpoint of PFS was significantly longer for the olaparib group (7.4 months) than the control group (3.6 months) with HR = 0.34, 95% CI, 0.25–0.47; p < .001. Secondary outcomes were more mixed. ORR favored the olaparib group (33%) versus the control group (2%) with OR = 20.86, 95% CI, 4.18–379.18; p < .001. The olaparib group also had a longer time to pain progression than the control group (HR = 0.44, 95% CI, 0.22–0.91; p = .02). Although there was a trend in favor of olaparib (18.5 months) compared to control (15.1 months) in terms of OS, this result was not statistically significant. A reduction in PSA by at least 50% (PSA₅₀) (45% vs 8%) and clearance of circulating tumor cells (30% vs 11%) also trended in favor of olaparib versus control, respectively. Data for cohort A were pooled with that of Cohort B, which included patients with alterations in any of the other 12 of 15 HR-related genes (BRIP1, BARD1, CDK12, CHEK1, CHEK2, FANCL, PALB2, PPP2R2A, RAD51B, RAD51C, RAD51D, and RAD54L). The secondary endpoint of PFS in the overall population also favored olaparib (5.8 months) over control (3.5 months) (HR = 0.49, 95% CI, 0.38–0.63; p < .001).

Moreover, the PROfound trial stratified patients according to whether they previously received taxane treatment. Exploratory analysis showed that patients who previously received taxane therapy benefited from olaparib treatment, and even those who did not receive taxane therapy had a trend toward benefit with olaparib treatment. In contrast, patients in both TOPARP-A and TOPARP-B must have received one or two prior chemotherapy regimens by design.^16,20 An obvious avenue for future research is to further define the role and outcomes of olaparib therapy after chemotherapy.

This study yielded safety data that were similar to previous trials. The median total duration of treatment for the overall population (cohorts A and B) favored the olaparib group (7.4 months) versus control group (3.9 months). However, the incidence of Grade 3 or higher adverse events (AEs) was more frequent with olaparib treatment than with control treatment. The main AEs of any grade were anemia, nausea, fatigue, or asthenia with olaparib, and fatigue or asthenia with control therapy. Of note, patients treated with olaparib had higher incidence of pulmonary embolism (4%) compared to control (1%), though the absolute numbers were small and none of the cases were fatal. There were no reported cases of myelodysplastic syndrome or acute myeloid leukemia in the PROfound trial. Moreover, the discontinuation rate of olaparib is 18% in the PROfound trial, which is significantly higher than the 6% observed in TOPARP-A,¹⁶ but was more in line of the 19% seen in TOPARP-B.²⁰

Caveats of olaparib therapy

One notable caveat is that the PROfound trial identified patients with HR-related aberrations using an investigational assay that is based on FoundationOne CDx next-generation sequencing. This assay was performed on biopsy tissue from primary or metastatic disease. To our knowledge, this test is not currently available on the market, and whether the results of the PROfound trial may be applied to patients identified as having aberrations in HR-related genes using other assays is unknown. Predictably, oncologists in the community will most likely identify somatic aberrations using a myriad of different assays, so these data certainly need to be validated for alternative assay types. In addition, it is also unclear whether there are inherent biological differences between patients who have germline versus somatic mutations in HR-related genes. Whether the results of the PROfound trial can also be applied to patients with germline mutations in HR-related genes remains to be demonstrated.

Another concern of this study is that there were key differences in baseline characteristics of the patients in the olaparib and standard-therapy groups. The standard-therapy group had higher baseline PSA levels, more patients with visceral metastases, and an ECOG score of 1. This represents a sicker population and could potentially explain why standard therapy was not as effective compared to olaparib. In contrast, patients in the olaparib group had more baseline alterations in ATM. Subgroup analysis also showed that patients with ATM did not derive as much benefit from olaparib. Therefore, this factor may have underestimated the benefit of olaparib compared to standard care. Given the opposing effects of these baseline factors, we suspect that the results presented likely indicate the true efficacy of olaparib.

An important criticism of the PROfound trial is that all patients previously received enzalutamide or abiraterone, and by design, the control treatment also consisted of either enzalutamide or abiraterone. If the patient’s disease has already become resistant to either enzalutamide or abiraterone, then subsequent treatment targeting the androgen receptor axis is likely to be less effective unless the mechanism of resistance is overcome. Repeating enzalutamide or abiraterone as the control treatment in this trial without knowing underlying mechanisms of treatment resistance may have diluted the efficacy of the control group, which could lead to biased outcomes that favor olaparib. Nonetheless, this strategy is consistent with the standard of care and is a reasonable design. Although olaparib demonstrated a trend of benefit in terms of overall survival for patients with BRCA1, BRCA2, and ATM, this outcome was not statistically significant. The authors attributed this result to the high rate of crossover of more than 80%, which likely diluted the efficacy data.

Conclusion

This trial will shape clinical practice in one of two ways. First, it verifies that mutations in genes related to HR, especially aberrations in BRCA1, BRCA2, and ATM, are not merely bystander mutations, but are lynchpins that can be strategically targeted for antineoplastic therapy. These findings will pave the way for additional testing – either germline or somatic – to guide therapy. It must be noted that various assays available on the market need to be compared in this patient population in order to derive the most benefit for these patients. Second, this trial definitively shows the efficacy of olaparib versus standard therapy for patients with mutations in HR-related genes and demonstrates that olaparib is an important addition to the armament of agents against metastatic prostate cancer. Olaparib seems to be especially effective as second-line therapy. Current research is focused on bringing olaparib to the front-line setting for metastatic, castration-resistant disease;²¹ as neoadjuvant therapy prior to prostatectomy;²² and as therapy for biochemically-recurrent disease after prostatectomy.²³ Moreover, active research in evaluating combination therapy that includes olaparib is currently being undertaken. For instance, preliminary results of combining olaparib with pembrolizumab after previous treatment with docetaxel seems promising.²⁴ Of note, another PARP inhibitor, rucaparib, has also shown efficacy in prostate cancer patients with alterations in DNA damage repair genes,²⁵ and has recently gained accelerated approval for those with BRCA mutations who have already been treated with chemotherapy.²⁶ It remains to be demonstrated how olaparib compares to other PARP inhibitors in this population of patients, and whether their differences, especially regarding toxicities, can be exploited to minimize dose-reductions and avoid discontinuations in order to maximize the benefits of therapy.

Disclosure of potential conflicts of interest

No potential conflicts of interest were disclosed.

Additional information

Funding

This work was supported by the National Cancer Institute [ZIA BC 010547].