Pinpointing the Need for Genetic Testing in Oncology

In Partnership With:

Partner | Cancer Centers | <b>UPMC Hillman Cancer Center</b>

Phuong L. Mai, MD, discusses the situations in which genetic testing should be considered and the importance of bridging awareness and access within oncology.

Referral for genetic testing for patients diagnosed with ovarian, pancreatic, and prostate cancer is a must, said Phuong L. Mai, MD, who added that referrals for other cancers, such as breast cancer, can depend on a number of factors, such as age at diagnosis, ethnicity, and background. However, the one constant that indicates a need for genetic testing is the variant allele fraction on somatic testing, she stated.

“We do see more variants on tumor testing that might be suggestive of a germline variant rather than a somatic variant. It’s not obvious on our test report, but the variant allele fraction certainly is something very important to consider. Above a certain variant allele fraction, the likelihood of that variant being germline is higher,” said Mai.

In an interview with OncLive® during the 2020 Institutional Perspectives in Cancer webinar on precision medicine, Mai, a hematologist/oncologist at the University of Pittsburgh Medical Center Hillman Cancer Center, discussed the situations in which genetic testing should be considered and the importance of bridging awareness and access within oncology.

OncLive: When should patients be considered for genetic testing?

Mai: It depends. There are some more straightforward indications for consideration for referral for genetic evaluation and perhaps testing. Some of the other indications could be a little complicated. Awareness is the key. Be aware of when consideration should be made for referral. If the patient doesn’t need to be tested or doesn’t want to be tested, that’s fine, but the referral would be the first step.

Mostly, we make the determination based on either personal cancer history or family history. There are some guidelines out there that one could follow. Mostly, we use the National Comprehensive Cancer Network [(NCCN) guidelines], but the United States Preventive Services Task Force [USPSTF] does have a set of guidelines as well. [Genetic testing] should be considered when we see an individual with a cancer diagnosis, such as ovarian cancer or pancreatic cancer. For an early diagnosis of breast cancer, [younger than 45 years old] is normally what we go by. Or, sometimes [patients] have a family history suggestive of an underlying genetic predisposition. In that case, we most certainly would suggest that [testing] be considered.

What are some of the most prevalent genetic abnormalities in breast, ovarian, and colorectal cancers (CRC)?

In terms of genetic predisposition to certain cancer diagnoses, BRCA or the hereditary breast/ ovarian cancer syndrome related to pathogenic variants in BRCA1/2 genes are the more common predisposition conditions. We tend to hear about [BRCA] a little more as well because we’ve known about it for quite some time. [BRCA] is mostly related to breast cancer and ovarian cancer risk. Lynch syndrome is another condition that’s relatively more common after rare genetic predispositions, and that’s mostly related to colon cancer risk and uterine cancer risk—mostly endometrial—and ovarian cancer risk to some extent. Those are the conditions that we tend to test for a little more often.

What are some of the key NCCN guidelines with regard to genetic testing in cancers of the breast, ovarian, pancreas, and prostate?

The main NCCN guidelines are referral and consideration for testing for all patients diagnosed with ovarian cancer, pancreatic cancer, and metastatic prostate cancer, or prostate cancer diagnosis, [because these patients] are considered to be at high risk. Those are in and of themselves indications for consideration for testing. In terms of a breast cancer diagnosis, most [patients with] breast cancer diagnosed before the age of 45, or triple-negative breast cancer diagnosed before the age of 60, [should undergo testing].

[Testing guidelines take into account] family histories [as well]. There are many different criteria in terms of family history. It’s a little harder to keep track of them all. But the bottom line is, [patients with] breast cancer diagnosed early [or with] a family history [should undergo testing]. Ethnicity and background do play a role. Individuals of Ashkenazi Jewish [ancestry] have a higher prevalence of having a variant in BRCA1/2. For individuals of that background, breast cancer diagnosis alone is an indication to consider testing.

CRC diagnosed before the age of 50 years, and family history of certain cancer diagnoses that are related to Lynch syndrome, such as endometrial cancer [would be an indication for testing]. In terms of CRC and Lynch syndrome testing, there is universal tumor testing to indicate whether or not Lynch syndrome testing should be pursued based on microsatellite instability or immunohistochemistry. If those findings are indicated on the tumor, certainly genetic testing should be considered. Otherwise, it’s a combination of personal and family history.

Do the NCCN and USPSTF guidelines differ at all?

They’re slightly different. The USPSTF guidelines tend to consider more of a population utility compared with those of the NCCN, where I think the focus is slightly more personal utility.

What considerations do you make with regard to referring patients for genetic testing based on the results of their tumor testing?

We do see more variants on tumor testing that might be suggestive of a germline variant rather than a somatic variant. It’s not obvious on our test report, but the variant allele fraction certainly is something very important to consider. Above a certain variant allele fraction, the likelihood of that variant being germline is higher. We normally use a cutoff of 30%, but that’s pretty rough. Some genes are more. We do see somatic changes in some genes a lot more than others. If we see a variant in a gene frequently somatically mutated in a tumor, it’s perhaps a lot less concerning than if we see a variant in a gene that is uncommonly somatically mutated in a tumor. Certainly, that should [warrant] consideration [for genetic testing]. [My recommendation] depends on a number of things. It depends on the tumor itself. It depends on the variant allele fraction and the gene itself. I don’t think there are strict rules that you can go by but be on the lookout for certain factors.

How does the confirmatory rate influence the role of genetic testing?

Some genes are more likely to represent a germline variant rather than a somatic variant if you were to see it in the tumor, and some are less so. Again, the allele fraction does matter. That’s the bottom line. One needs to be aware of the possibility that certain findings on tumor testing could represent a germline variant, and a patient can certainly benefit from confirmatory testing for that variant. However, not everybody with a variant on their tumor would need to be confirmed, because the majority of them would be somatic changes rather than a germline change.

What challenges remain with regard to genetic testing?

The main challenges have to do with access to care or access to genetic services. Genetic cancer services are still not widely available. Certainly, in big cities, you have access to those clinics. However, a rural area might not be so. Traveling 3 or 4 hours to a genetic center may or may not be feasible for everyone.

One of the other barriers that we have seen has to do with awareness, and that has to do with providers and patients alike. Some providers don’t have their radars on in terms of recognizing indications for referral to genetic services. Also, some patients are somewhat skeptical of the whole thing in terms of genes and genetic testing and family history for that matter. Culturally, there are some barriers as well. In some cultures, speaking about cancer diagnosis in a family might not be something that is done routinely. One of the other barriers is perhaps financial, right? This is because testing might not be covered, and it could be cost prohibitive for some individuals.