Christopher Wallis: Hello, and thank you for joining us for this UroToday discussion of the NCCN clinical practice guidelines in oncology with a focus on the prostate cancer guidelines from February, 2021. Today, we're discussing prostate cancer genetics. I'm Chris Wallis, an assistant professor in the Division of Urology at the University of Toronto, and with me is Zach Klaassen, an assistant professor in the Division of Urology at the Medical College of Georgia.

In terms of discussing prostate cancer genetics, we'll go through a little bit of background first and then focus on three specific aspects of prostate cancer genetics, namely homologous recombination repair, DNA mismatch repair, intraductal/cribriform or ductal histology, and then we'll finish off by discussing the NCCN genetic testing recommendations.

In terms of background, most will know that prostate cancer is highly heritable, despite the fact that specific predisposition genes are not a highly prevalent. So family history clearly raises the risk of prostate cancer, and some of these patients will have known syndromes, like hereditary breast and ovarian cancer, Lynch syndrome, or others, but the vast majority of patients with a positive family history won't have one of these known predisposition syndromes, and instead will simply have an elevated risk that is heritable. And so, as a result, the panel recommends a thorough review of both personal history for other malignancies and family history for all patients with prostate cancer.

In the context of genetic testing, we need to consider both germline and somatic mutations. Germline are those that are present in every cell in the body and have been passed down through generations. The importance of germline mutations are multiple. First, when germline mutations are identified, family genetic counseling is necessary for cascade testing, as this may affect relatives and offspring. Further, germline mutations are associated with these cancer risks syndromes, as we alluded to before, and may be important for assessment of the personal risk of subsequent or other related malignancies. The best example comes from women with a BRCA1/2, who are at risk of multiple malignancies, including breast and ovarian cancers. Finally, germline mutations, when identified, may have important therapeutic implications.

In terms of somatic mutations, these are ones that occur in the tissues after birth, and so these, in the context of carcinogenesis, are typically restricted to the tumor. These are frequently found in prostate cancer in the absence of germline mutations, and so some studies have shown that almost 90% of patients with MCRPC have potentially actionable somatic mutations, and only 10% of these correspond to an underlying germline mutation.

First, we're going to talk about DNA repair genes. In terms of alterations in these genes, mutations are found in nearly 1 in 5 patients with localized prostate cancer and 1 in 4 with metastatic disease. Most commonly, these affect BRCA2 and ATM, although BRCA1 and numerous other genes highlighted in this schematic here, including RAD51 and PALB2, are implicated and may be affected. As we sort of alluded to before, the prevalence of germline DNA repair mutations increases with advancing and higher-risk disease characteristics. So in patients with lower-risk localized disease, they occur current 1.5% to 4%, in those with higher-risk localized disease, 6% to 9%, and those with metastatic disease, 7% to 16%. One of the more seminal papers on this topic from Dr. Pritchard and colleagues found that nearly 12% of men with the MCRPC had germline mutations in one of 16 DNA repair genes. Notably, a family history of breast cancer increases the likelihood of identifying germline DNA repair mutations, with an odds ratio of nearly 1.

And so this study, unlike Pritchard's work, looked at a more heterogenous cohort, and they looked at a variety of different germline mutations, not just homologous recombination repair. But they found positive germline variants in 17%, and 30% of all variants were BRCA1 or BRCA2, which are implicated in homologous recombination repair.

It's worthwhile noting that these kinds of mutations may be enriched in certain populations, in particular, among Ashkenazi Jews. The rate of BRCA1 and BRCA2 mutations in Ashkenazi populations exceeds 2% and carriers have a 16% chance, that's 1 in 6, of prostate cancer by age 70. Further, even among Ashkenazi men, BRCA1 and BRCA2 mutations are associated with higher rates of prostate cancer.

Just to dive into BRCA1 and BRCA2 a little bit more as the guideline panel did, mutations in these genes increased the risk of prostate cancer, in particular, BRCA2 may be associate with a two to six-fold increase in prostate cancer risk, and even beyond this, not just increasing the incidence of prostate cancer, BRCA2 may also be associated with an earlier age of diagnosis, more aggressive phenotype, and worse survival. However, these mutations that do have important therapeutic implications, which we've discussed in previous talks.

At this point in time, I'm going to pass it over to Zach to walk us through a discussion of a DNA mismatch repair and intraductal and cribriform histology.

Zachary Klaassen: Thanks, Chris. We'll start off by discussing the in MLH, MS2, and MSH6, as well as PMS2, which may confer microsatellite instability and deficient mismatch repair. Importantly, these are associated with germline mutations as well as associated with Lynch syndrome, which we'll discuss in more detail. Lynch syndrome was also known as hereditary nonpolyposis colorectal cancer, and you can see that here, that there's two iterations of criteria that make up Lynch syndrome.

First, there was the Amsterdam Criteria 1, which stated that there should be at least three relatives with colorectal cancer, all with the following criteria, including one first-degree relative of the other two relatives, at least two successive generations should be affected, at least one colorectal cancer should be diagnosed before the age of 50, familial adenomatous polyposis should be excluded, and tumors should be verified by pathological examination.

Subsequently, the updated revised Amsterdam Criteria 2 included there should be at least three relatives with a hereditary nonpolyposis colorectal cancer, and this includes colorectal cancer, cancer of the endometrium, small bowel, ureter, or renal pelvis. Similar to the Amsterdam Criteria 1, we see that one should be a first-degree relative of the other two relatives, at least two successive generations, at least one cancer diagnosed before the age of 50, familial adenomatous polyposis should be excluded, and tumors should be verified by pathological exam.

When we look at the ICG definition of Lynch syndrome, we see that this includes familial clustering of colorectal and endometrial cancer with other associated cancers, spreading a wide variety, including stomach, ovarian, upper tract, brain, small bowel, hepatobiliary, and skin. These cancers develop at an early age. Multiple cancers can develop. And you can see several features here of colorectal cancer. These tumors are also associated with a high frequency of MSI-high testing, as well as immunohistochemistry, including loss of MLH1, MSH2, or MSH6 protein expression, and germline mutation in the MMR gene, such as MSH2, MLH1, MSH6, PMS1, and PMS2. Notably, as we see here on the bottom left, in the revised Amsterdam Criteria as well as in the Lynch syndrome, this criteria does not require a family history of colorectal cancer, but focuses on the constellation of cancers that may be involved with Lynch syndrome.

With regards to Lynch syndrome and prostate cancer, prostate cancer is not one of the classic Lynch-associated cancers. However, patients with Lynch may have an increased risk of prostate cancer. Notably, previous studies have also shown that older patients with germline MSH2 mutations may have a higher risk of prostate cancer. This study in JCO published in 2019 looked at the rate of MSI-high in a variety of cancers associated Lynch syndrome. This study found that MSI-high was in 5% of patients with prostate cancer, including 5.6 of those that had Lynch syndrome. So overall, about 2.29%. This corroborates other studies that have shown rates of between 1.7% to 3.1%.

Next, we'll talk about the effect of intraductal and cribriform or ductal histology on prostate cancer. Ductal prostate cancer is rare. It makes up about 1.3% of prostate cancer, with intraductal prostate cancer being more common and enriched among those with high-risk prostate cancer. Between these two, there's significant overlap of diagnostic criteria and intraductal, ductal, and invasive cribriform patterns may co-exist in the same patient.

A little more about intraductal and cribriform. There's limited data on the association between ductal or intraductal histology and increased genomic instability, however, the studies that have looked at this have shown an increased likelihood of MMR gene alterations, as well as an increased likelihood of germline homologous DNA repair gene mutations. Patients with germline BRCA2 mutations do commonly have intraductal histology, and the overall, the panel believes that data connecting histology and genomic alterations are stronger for intraductal rather than ductal histology. Furthermore, patients with intraductal carcinoma on biopsy should have germline testing according to the panel.

Finally, we'll talk about genetic testing recommendations. Germline genetic testing is recommended for patients with prostate cancer and any one of the following features listed here, including family history, high-risk or very-high-risk, regional or metastatic prostate cancer, those with an Ashkenazi Jewish ancestry, or those patients with intraductal cribriform histology. When testing for these patients, this should include several markers or several gene alterations, including ML1 MSH2, MSH6, PMS6, BRCA1 and 2, ATM, and CHEK2, and other genes may be appropriate to test as well, including HOXB13.

Finally, cancer predisposition next-generation sequencing can also be considered for these patients. With regards to germline testing, genetic counseling and support is critical for these patients, and pre-test counseling is recommended, particularly in those of the family history. Post-test counseling for those identified pathologic variants is important, as well as consideration for cascade testing for those with pathogenic variants.

Testing for somatic homologous recombination mutations should be considered for those with regional disease and is recommended for those with metastatic disease. Testing for MSI and dMMR is considered in those with regional or metastatic castrate-sensitive prostate cancer, and is definitely recommended for those metastatic castration-resistant prostate cancer. Multigene molecular testing should be considered for patients with low, intermediate, and high-risk localized disease, and those with a life expectancy of greater than 10 years, and the panel specifically mentioned the Decipher molecular assay that should be considered for patients with adverse features following radical prostatectomy.

A few more notes on somatic testing. If somatic testing identifies mutations in the homologous recombination genes, patients should be referred for genetic counseling, and if somatic testing identifies MSI or dMMR mutations, patients should be referred for genetic testing to assess for the possibility of Lynch syndrome, which may have implications for future therapeutics, including eligibility for pembrolizumab. The guidelines recommend that over interpretation of germline findings should be avoided on the basis of somatic tumor testing and sequencing.

Additional testing has evolved over the last several years, including the evaluation of the antigen receptor pathway and patients with MCRPC. AR-V7 testing, which can be based on a circulating tumor cells, may allow for systemic therapy. And in this paper from JAMA oncology in 2016, we see that those that were AR-V7 positive had a significant benefit receiving a taxane versus androgen receptor inhibitor, whereas those that were AE-V7 negative did not see the same benefit for taxane chemotherapy.

In conclusion, for this prostate cancer genetic section, we see that for homologous DNA repair genes, these are relatively common, with the rates increasing along the natural history of prostate cancer, as well as being more common in Ashkenazi Jewish populations. There's an increased risk of prostate cancer and worse phenotype in patients that do have these mutations, and in previous discussions we've had, we note that these mutations do have treatment implications, including eligibility for receipt of PARP inhibitors. In terms of DNA mismatch repair genes, these are associated with Lynch syndrome, which is relatively uncommon in prostate cancer, less than 5%, but may have treatment implications in terms of eligibility for pembrolizumab. With regards to intraductal and cribriform and ductal prostate cancer, this is associated with higher risk disease, as well as associated with genomic alterations. When thinking about testing, we should consider patients with family history, their histology, and their disease characteristics by risk group.

Thank you very much. We hope you enjoyed this NCCN and prostate cancer panel guide discussion, looking specifically at prostate cancer genetics.