CME
Physicians: Maximum of 1.00 AMA PRA Category 1 Credit™
Released: March 22, 2024
Expiration: March 21, 2025
Prostate Cancer May Be Diagnosed at Various Stages and Progress Through Different Pathways
Himisha Beltran, MD:
It is a pleasure to be here with pathologist Dr Steven Smith to discuss the rapidly evolving data and guidance on biomarker testing and its implications for treatment decisions in prostate cancer. I will provide a quick introduction to this disease from my perspective as a medical oncologist.
Prostate cancer is the most common cancer among men in the United States and is expected to cause 35,250 deaths in 2024, unfortunately.1 As shown in the schematic here, this heterogenous disease can be diagnosed at different stages and progress through various pathways. A minority of patients present with de novo metastatic prostate cancer, with most presenting with localized disease detected through PSA screening.2
After undergoing definitive treatment with curative intent for localized prostate cancer, a subset of patients will experience recurrence.3,4 This is usually “biochemical recurrence”—defined as an increase in blood levels of PSA without evidence of metastasis on conventional imaging—that eventually progresses to metastatic disease.
Both de novo (“synchronous”) and recurrent (“metachronous”) metastatic hormone‑sensitive prostate cancer (HSPC) are treated primarily with doublet therapy comprising ADT in combination with an AR pathway inhibitor. Depending on the volume of the disease and the clinical context, radiation therapy and/or chemotherapy may be added.
Despite advances in management of metastatic HSPC, our treatments are not curative, and patients do progress to CRPC—disease that keeps growing despite castrate levels of testosterone.5 Because most deaths occur in those with metastatic CRPC, this disease setting has been a major focus for research and drug development.6 Metastatic CRPC will be our focus today as we discuss advances in biomarker identification and biomarker‑directed therapy with PARP inhibitors.
Targeting DNA Repair in Prostate Cancer: A Note on Nomenclature
Steven Christopher Smith, MD, PhD:
Thank you, Dr Beltran, for that background on prostate cancer. Before we start on our discussion about biomarker testing, it is important to briefly review the relevant nomenclature.
DDR refers to DNA damage repair, and we will shortly review the different pathways used by cells to repair various types of DNA damage.7-9
DRD refers to DNA repair deficiency in cancer cells resulting from an impairment of at least one of the DDR pathways.
HR refers to homologous recombination, which refers to a key, high fidelity part of the function of a family of genes involved in a DDR pathway—specifically, the HRR pathway for repair of double-strand breaks of DNA.
HRR refers to homologous recombination repair and is synonymous with HDR (homology‑directed repair). This DDR pathway involves numerous genes that will be our focus today as we discuss biomarker testing and biomarker-directed therapy.
HRD, or homologous recombination deficiency, is a specific type of DRD. Cancers with HRD are deficient in function of the HRR pathway.
DNA Repair Pathways
Steven Christopher Smith, MD, PhD:
There are numerous DNA repair pathways, from mismatch repair—which is involved in Lynch syndrome10—to those involved in repairing double-strand breaks, which will be our focus. Double-strand breaks can be repaired through the non-HR pathway, which is error prone, or through the HRR pathway, which uses the homologous strand of DNA as a template for high-fidelity repair.11
However, the HRR pathway is not the direct target of PARP inhibitors. These agents inhibit the base excision repair pathway, which is involved in repairing single-strand breaks in DNA. In base excision repair, PARP1 attracts repair proteins and cofactors to the site of DNA damage, where the altered bases are removed, and the single strand break is repaired using the other DNA strand as a template. Once repair is complete, PARP1 detaches from the DNA duplex via self-glycosylation.
From BER to HRR: PARP Inhibition in Cell With HRR Function
Steven Christopher Smith, MD, PhD:
How do we get from base excision repair to homologous recombination repair? The connection is through inhibition of 2 enzymes vital to base excision repair: PARP1 and PARP2.11
Inhibition of PARP1 in the base excision repair pathway leads to accumulation of single-strand breaks. Inhibitors also “trap” PARP on the DNA during cell replication, thereby leading to conversion of single-strand breaks to double-strand breaks. Thus, cells experiencing PARP inhibition become very reliant on the HRR pathway to repair these double-strand breaks to avoid having to use the error-prone non-HR pathway, which frequently generates mutations.
From BER to HRR: PARP Inhibition in Cell With HRR Deficiency
Steven Christopher Smith, MD, PhD:
But what if that cell also has HR deficiency, such as a BRCA mutation, when PARP1 and PARP2 are inhibited? PARP inhibition results in unrepaired single-strand breaks, which are converted into double-strand breaks that cannot be repaired by the deficient HRR pathway, forcing cells to rely on the error-prone non-HR pathway. This leads to accumulation of DNA damage and ultimately cell death. This phenomenon is known as “synthetic lethality,” in which loss of function of both PARP1/2 and the HRR pathway results in cell death, but loss of function of either alone does not.
Synthetic Lethality With PARP Inhibition Is Produced by Conditional Drug Sensitivity in HRR-Deficient Cells
Steven Christopher Smith, MD, PhD:
This diagram illustrates the conditions necessary for synthetic lethality with PARP inhibitors. Normal cells are capable of both homologous recombination through the HRR pathway and single-strand DNA repair through the base excision repair pathway and are therefore viable.12-14 If normal cells have the base excision repair pathway blocked with a PARP inhibitor, the cells are still viable because the HRR pathway remains intact to rescue for it.
The opposite scenario is also true, where cells deficient only in HRR are still able to repair single-strand breaks through the base excision repair pathway and remain viable, despite lacking high-fidelity HRR. However, blocking the base excision repair pathway with a PARP inhibitor in HRR-deficient cells induces synthetic lethality where these cells ultimately die due to accumulated DNA damage.
“Synthetic lethality” provides a very promising angle for drug development because pharmaceutically inhibiting the function of the one gene or pathway causes death of tumor cells only when the other target gene is already mutated. This limits toxicity to normal cells without the relevant synthetic lethal mutation. In metastatic CRPC, our target is HRD via one of several genes in this pathway, with PARP inhibition representing the partner that, together with HRD, is synthetically lethal.
HRR Genes With Approved PARP Inhibitor Indications
Steven Christopher Smith, MD, PhD:
There are numerous genes encoding proteins that participate directly or indirectly in the HRR pathway. Mutations in these genes—of note, BRCA1 and BRCA2—may lead to HRD and vulnerability to synthetic lethality when the cell is exposed to PARP inhibitors.15 That being said, the sensitivity to PARP inhibitors varies by gene, as Dr Beltran will discuss later.
Currently, the FDA has approved at least 1 PARP inhibitor–based therapy for patients with metastatic CRPC harboring alterations in ≥1 of the following HRR genes.16-21
HRR Mutations Occur in 25% of Patients With Metastatic CRPC
Steven Christopher Smith, MD, PhD:
DNA damage repair alterations, primarily in the HRR pathway, are quite common in prostate cancer, occurring in 20% to 30% of patients with metastatic CRPC.22-25 BRCA2 is the most commonly mutated HRR gene, with tumor BRCA2 alterations reported in 8.7% of the metastatic CRPC population in the PROfound trial and germline BRCA2 alterations observed in 3.5% to 5.3% of patients across stages of prostate cancer. This highlights the importance of somatic tumor testing—if only germline testing is performed, up to one half of patients with HRR mutations may not be detected.
HRR Pathways Can Be Deficient in Germline DNA, Tumor DNA, or Both
Steven Christopher Smith, MD, PhD:
Germline mutations in HRR genes are passed from parents to offspring by autosomal dominant inheritance.26 In the depicted scenario where the father has a germline HRR mutation but the mother does not, approximately one half of their offspring will obtain the normal copy and one half the germline deleterious variant from the affected father based on Mendelian inheritance patterns. Those who inherit a dysfunctional copy have a significantly elevated risk of prostate and other cancers, especially for BRCA2, because it takes only a second somatic mutation to contribute to oncogenesis. By contrast, an individual without an inherited germline mutation must experience 2 somatic mutation events to exhibit the same HRR deficiency.
Thus, it is not surprising that among patients with metastatic prostate cancer who harbor an inherited germline HRR mutation, approximately 60% exhibit a somatic alteration in the second allele.23 Furthermore, patients with advanced prostate cancer are more likely to harbor mutations in HRR genes than those with earlier disease.
Multiple Paths for Biomarker Testing in Prostate Cancer
Steven Christopher Smith, MD, PhD:
Now that we have discussed the rationale for using PARP inhibitors in prostate cancers harboring HRR mutations, we can move on to discussing how to test for those mutations. First, however, we should consider why patients would choose to undergo biomarker testing and how the patient’s goals affect the type of testing.
Patients undergo testing for 2 general reasons. The first reason is to determine whether the patient has an actionable HRR mutation and is thus eligible for PARP inhibitor–based therapy. Somatic and/or germline testing can be used for determining eligibility.
The second reason is to determine the risk for developing various cancers in the patient and other family members. For example, BRCA2 mutations are associated with a significantly increased risk in developing prostate cancer.27 Germline testing is used for assessing risk and determining potential relevance to family members.
Sample Sources for Biomarker Testing: Tumor vs Liquid Biopsy
Steven Christopher Smith, MD, PhD:
An important consideration is what sample will be used for biomarker testing. The gold standard continues to be a tumor tissue biopsy, either of primary tumor or a metastatic lesion.28
However, healthcare professionals (HCPs) are increasingly turning to a blood-based “liquid” biopsy. A liquid biopsy takes advantage of tumors shedding DNA into systemic circulation—hence “circulating tumor DNA” (ctDNA). Although normal cells in the body also shed DNA, individuals with cancer have a substantially greater burden of circulating DNA. The ctDNA also tends to be in smaller fragments than normal circulating DNA.29
This table summarizes some of the important advantages and challenges when using a tumor biopsy vs liquid biopsy.5,30-32
The advantages for tumor biopsy include that this provides tumor‑enriched samples that can be assessed for purity. Tumor biopsy also enables detecting copy number changes (eg, the zygosity of mutations, homozygous loss) and measuring tumor mutational burden (TMB), which determines eligibility for immune checkpoint inhibitor–based therapy. Another important advantage of tumor biopsy is the ability to exclude neuroendocrine transformation. Although less common, neuroendocrine transformation can happen during the progression of prostate cancer and requires different therapeutic strategies.
Regarding disadvantages, biopsying a primary tumor or metastasis is certainly more involved than venipuncture for liquid biopsy. This also makes it more challenging to do serial sampling during the disease course. There is always the potential for failed samples with inadequate tumor tissue, necrosis, or other factors. Furthermore, tumor biopsies can be heterogenous both within and across metastatic sites, as well as over time.
The advantages for liquid biopsy include being able to detect somatic mutations with high fidelity. Liquid biopsy is certainly more convenient and less invasive than a tumor biopsy. Liquid biopsy has a low testing failure rate—approximately 3%—in metastatic CRPC. Furthermore, liquid biopsy permits monitoring over time with serial samples.
That being said, the challenges of liquid biopsy include that ctDNA can be quite fragmented. Thus, this approach can miss certain types of alterations, especially deletions, and yield a false-negative result. Furthermore, ctDNA yield depends on tumor burden. There is also the potential for interference due to a phenomenon unrelated to prostate cancer known as clonal hematopoiesis of indeterminant potential (ie, CHIP).
When to Test: Prostate Cancer Disease States
Steven Christopher Smith, MD, PhD:
The timeline here suggests 3 key points where testing has implications for family counseling and/or treatment decisions for the patient. The first is at the diagnosis of clinically localized disease, where identification of germline HRR mutations has immediate implications for family counseling, especially for those with a strong family history or with high-risk cancer.33 The other key points are during the transition to metastatic disease and at secondary progression, where identification of HRR mutations can make a patient eligible for PARP inhibitor–based therapy.
Dr Beltran, when do you perform testing in your patients with prostate cancer?
Himisha Beltran, MD:
There is no perfect consensus on the optimal time to test. In my practice, I recommend germline testing at the time I meet a patient with high‑risk or metastatic prostate cancer, or those with a strong family history, where there is a higher likelihood of having a germline alteration with important family risk implications.
For somatic testing, I usually start thinking about the testing at the time of metastatic disease, even hormone‑sensitive disease, because it can take time to locate an archival prostate biopsy or prostatectomy samples. Because HRR mutations are typically early events, primary prostate cancer samples have been used in most clinical trials testing PARP inhibitors.34 That said, in addition to challenges in locating these older samples, which may be stored at an outside clinic, there can be a high failure rate with archival primary samples.
When it is not feasible or possible to perform testing on primary samples before CRPC, I wait until the patient progresses and then either do ctDNA testing or consider a metastatic biopsy, recognizing the limitations that you mentioned. If a patient has soft tissue or visceral disease, I have a low threshold to do a metastatic biopsy. Bone biopsies are more challenging to do, so I do not perform bone biopsies in every single patient at the time of CRPC. ctDNA can be useful in these cases and has demostrated high concordance with biopsies when it comes to common DNA repair mutations, though is dependent on ctDNA fraction and is less sensitive for deletions.
As far as serial testing, we do not currently have data supporting the utility and cost effectiveness of doing ctDNA testing at every progression. But in some situations—like if I had a negative ctDNA test in the context of a very low tumor fraction—then I would repeat testing at another progression. As you mentioned, there must be disease burden to be able to detect ctDNA, so we do not do ctDNA testing when patients are responding to treatment. We must wait until their PSA is rising and/or if there are new developments radiographically.
Steven Christopher Smith, MD, PhD:
Do you test any earlier if the patient has high-risk disease?
Himisha Beltran, MD:
I personally do not do somatic testing in patients with localized prostate cancer unless we are considering a clinical trial. My reasoning is that if the patient never recurs, then I will not be recommending PARP inhibitor–based therapy based on current indications.
Dr Smith, what is your perspective on testing in earlier disease?
Steven Christopher Smith, MD, PhD:
I want to note that the NCCN recommends testing for patients with high-risk or very high–risk localized disease and consideration of testing for patients with intermediate-risk prostate cancer with intraductal/cribriform histology, even if localized.35 However, there is some controversy with this latter recommendation. Although intraductal/cribriform prostate cancer is enriched for HRR mutations,36 it is unclear whether these mutations are independently related to histology or the relationship is because this histology is prevalent in higher‑grade disease.
Current Guideline Recommendations for Somatic (Tumor) Testing
Steven Christopher Smith, MD, PhD:
Whenever I am participating in tumor boards and speaking to pathologists in the community, I like to emphasize that the standard of care for testing has been evolving very quickly. When we look for guidance on who should receive somatic (tumor) testing in prostate cancer, NCCN guidelines currently recommend somatic testing for all patients with metastatic disease and consideration of testing for those with regional disease.37 This should be multigene tumor testing for HRR alterations in at least 8 genes, including ATM, BRCA1, BRCA2, CDK12, CHEK2, FANCA, PALB2, and RAD51D. Dr Beltran soon will discuss their relevance to potential treatment.
The NCCN also recommends tumor testing for microsatellite instability (MSI)–high and/or mismatch repair deficient status (dMMR) in metastatic CRPC. MSI-high/dMMR testing should be considered for regional disease or metastatic HSPC, and TMB testing can be considered in those with metastatic CRPC. Again, these biomarkers are for determining eligibility for regimens based on immune checkpoint inhibitors, not PARP inhibitors.
Current Guideline Recommendations for Germline Testing
Steven Christopher Smith, MD, PhD:
Germline testing (ie, genetic testing) is recommended when the findings would affect treatment choice, clinical trial options, risk of other cancers in the patient, and risk to other family members.37-39 It is important to note that the recommendations for germline multigene testing include genes on top of those recommended for somatic testing. One key gene is HOXB13, which is a non-DDR gene related to risk of prostate cancer.
To help identify the patients and/or families most likely to benefit from germline testing, national guidelines use a spectrum of risk, as illustrated on the right. Germline testing is recommended for all patients with high-risk, very high–risk, regional, or metastatic prostate cancer. However, germline testing also is recommended for anyone with a strong family history, personal history of breast cancer, or Ashkenazi Jewish ancestry.
Himisha Beltran, MD:
These developments in testing and targeted therapy really have changed how we manage patients with newly diagnosed prostate cancer. I think this guidance emphasizes the importance of taking a family history and discussing genetic risks with our patients.
There is a great deal of variability in how germline testing is performed. In my institution, we refer patients to genetic counselors, who will obtain the full family history and perform counseling and germline testing. If positive, the counselors then do “cascade” family testing and counseling.
Unfortunately, not enough genetic counselors are available, even as more patients are being tested. In many institutions, HCPs are increasingly performing pretest counseling up front, and it is important to remember these guidelines.
Steven Christopher Smith, MD, PhD:
I appreciate you saying that. Having studied and participated in the treatment of hereditary cancers over many years, I am often struck by how effective clinical geneticists are at soliciting family histories. It is quite striking how revealing that consultation can be. Fortunately, at our institution, the wait has shortened recently for patients to see a clinical geneticist.
Germline: Critical Importance of Cascade Testing
Steven Christopher Smith, MD, PhD:
Dr Beltran briefly mentioned cascade family testing for patients found to have positive germline alterations. These are not rare: Approximately 1 in 10 men with metastatic CRPC have an inherited HRR mutation.23,28 Both siblings and children have a 50/50 risk of inheriting that same germline mutation; knowing their status will help them make decisions about how to manage their risk of other cancers.
Counseling Patients About Genetic Testing: Considerations and Consent
Steven Christopher Smith, MD, PhD:
As Dr Beltran mentioned, HCPs are increasingly performing genetic counseling up front for our patients with prostate cancer. There are some important considerations—particularly regarding consent—when providing genetic counseling.35
First, a person who has prostate cancer and/or a family history of cancer suggests the possibility of an inherited syndrome.
Second, it is important for patients to realize that testing, even by contemporary modalities, is not perfect. Testing may be able to identify mutations that contributed to the development of prostate cancer and that could affect treatment options.
Third, if family members are tested and found to have the same germline mutation, there may be screening or even interventions to reduce the risk of cancer in affected family members.
My fourth point—which is very important to many patients—is that federal Genetic Information Nondiscrimination Act (ie, GINA) laws protect an individual’s health insurance and employment, should they have a germline mutation predisposing them to cancer.40 However, life insurance and long‑term care insurance are not federally protected, and this can be very relevant to siblings or children.
Fifth, the potential outcomes of these tests include identifying:
For all those reasons, consent for testing and follow-up genetic counseling is vital.
Germline vs Tumor (Somatic) Testing
Steven Christopher Smith, MD, PhD:
When considering the advantages and limitations of germline and tumor testing, bear in mind that samples for germline testing are much easier to obtain than those for tumor (somatic) testing.30-32 The germline sample can be saliva, blood, or sometimes even a mucosal smear, and in those with prostate cancer, it can be obtained at any point in the disease course. Again, germline testing can provide information about prognosis, eligibility for targeted therapy, and potential trials, as well as potentially life‑saving information for relatives. However, if you perform only germline testing, you may miss up to 50% of patients with somatic-only HRR mutations who may benefit from PARP inhibitor–based therapy.
Tumor (somatic) testing is comprehensive—meaning we can detect genetic mutations that occurred only in the tumor, as well as those present in the germline. However, generally, tumor testing cannot distinguish between somatic vs germline mutations. In most cases, somatic testing is performed on biopsy tissue or ctDNA in the blood, but timing and conditions of the sample—especially if it is archival—can affect the quality of data. Again, somatic testing may provide information about prognosis, therapy options, or trials, but it does not directly provide information on risk to relatives or risk of other cancers in the patient. That being said, the tumor content and purity of the sample can affect the sensitivity and specificity of somatic testing. Serial testing is very difficult, but generally speaking—and as Dr Beltran mentioned—many of the known actionable mutations in HRR pathway are mostly truncal and arise early in tumor genesis.34
Clinical Pearls on Testing for Community HCPs
Steven Christopher Smith, MD, PhD:
Many HCPs find ordering and interpreting tests quite challenging, particularly in community settings without tumor boards. To finish up this section, I would like to provide a few quick tips for reading a tumor testing report.41,42
First, know what genes are tested for by the assay. You will want to make sure that the assay covers the relevant genes recommended by current NCCN guidelines. Often these are listed comprehensively in an appendix.
Second, know what types of alterations are detected by the assay. Not every testing approach is equally sensitive or specific for different kinds of alterations, from deletions to rearrangements to fusions.
Third, be aware that MSI and TMB status are reported only by a subset of assays. Assays without enough gene targets may not be able to identify TMB with high fidelity.
Dr Beltran, do you have any final “clinical pearls” on testing for our colleagues?
Himisha Beltran, MD:
Yes. Please note that some commercial reports include “tiers” indicating the functional significance of detected alterations. I would stress that it is important not to act on a VUS.
When reviewing a report, it is also important to consider the pathogenicity of the alteration. Is the alteration germline or somatic? Is it a monoallelic or biallelic alteration? Clinical trial data suggest that PARP inhibitors are more active against alterations that are biallelic, including those with homozygous deletions.43
Finally, do reach out to your local molecular pathologist. Even if it is a commercial assay and they are not actually reading the report, they could probably help answer any questions you may have.
Steven Christopher Smith, MD, PhD:
Absolutely.
Back
