The ability to find and target specific biomarkers in the DNA of advanced cancers is rapidly changing options and outcomes for patients with locally advanced and metastatic solid tumors. This strategy is the basis for precision oncology, defined here as using predictive biomarkers from tumor and/or germline sequencing to guide therapies. This article focuses specifically on the use of DNA sequencing to find those biomarkers and provides guidance about which test is optimal in a specific situation, as well as interpretation of the results. We emphasize the identification of biomarkers that provide adult patients with advanced solid tumors access to therapies that would not be an option had sequencing not been performed and that have the potential for significant clinical benefit. The best approach is to have an expert team with experience in precision oncology to assist in the interpretation of results.
Which test?
Deciding what test of the array of assays available to use and which tissue to test can be overwhelming, and uncertainty may prevent oncology practitioners from ordering germline or somatic sequencing. For the purposes of this article, we will focus on DNA sequencing for inherited/germline alterations (including mutations, copy number changes, or fusions), which may inform treatment, or alterations that arise in the process of carcinogenesis and tumor evolution (somatic alterations in tumor DNA). This focus is not meant to exclude any specific test but to focus on DNA-based tests in patients with locally advanced or metastatic malignancy.
Germline Testing
Germline testing is the sequencing of inherited DNA in noncancerous cells to find alterations that may play a role in the development of cancers and are actionable in some cases. Germline alterations can inform therapeutic decisions, predict future cancer risk, and provide information that can help family members to better manage their risks of malignancy. Detailed discussions of the importance of germline testing to inform cancer surveillance, risk-reducing interventions, and the testing of relatives to determine who carries inherited alterations (cascade testing) is extremely important with several advantages and is covered in a number of excellent reviews elsewhere.1-3 Testing of germline DNA in patients with a metastatic malignancy can provide treatment options otherwise not available for patients, particularly for BRCA1/2 and Lynch syndrome–related cancers. Recent studies have shown that 10 to 15% of patients with advanced malignancies of many types have a pathogenic germline alteration.4,5
Germline DNA is usually acquired from peripheral blood, a buccal swab, or saliva collection and is therefore readily available. This is advantageous because it does not require a biopsy to identify relevant alterations. Germline testing is also less susceptible to the rare situations in which artifacts occur in formalin-fixed tissues and obscure relevant alterations.
The cost of germline testing varies, but most commercial vendor assays for germline testing are significantly less expensive than the cost of somatic testing. The disadvantages include the inability of germline testing to find any alterations that arise solely in tumor tissue and the smaller gene panels included in germline testing as compared to somatic testing panels. Other considerations relate to the inherited nature of pathogenic germline variants and its implications for family members that may affect the patient’s psychosocial health and potentially change the family dynamics.
Deciding who is appropriate for germline testing and when to perform the testing should be individualized to the patient’s wishes and disease status. Treatment planning may be less complicated if testing has been performed and germline status is known. In some cases urgent germline testing is indicated to inform pending procedures and/or surgical decisions for risk reduction, including more extensive tissue resection, such as the removal of additional organs or contralateral tissue. A minor point regarding germline testing is that the DNA of patients with hematologic malignancies may be difficult to sequence because of sample contamination by the circulating malignancy. For this reason, most laboratories will not accept peripheral blood or saliva samples for germline testing in patients with active hematologic malignancies; they often require DNA from another source such as fibroblasts from a skin biopsy or cells from a muscle biopsy. Germline testing is recommended for all patients with metastatic prostate cancer, as well as any patient with any stage of pancreatic cancer or ovarian cancer and patients with breast cancer diagnosed at age ≤ 45 years. More detailed criteria for who is appropriate for germline testing outside of these groups can be found in the appropriate National Comprehensive Cancer Network (NCCN) guidelines.6-8 In patients with some malignancies such as prostate and pancreatic cancer, approximately half of patients who have a BRCA-related cancer developed that malignancy because of a germline BRCA alteration.9-11 Testing germline DNA is therefore an easy way to quickly find almost half of all targetable alterations with a treatment approved by the US Food and Drug Administration (FDA) and at low cost, with the added benefit of providing critical information for families who may be unaware that members carry a relevant pathogenic germline alteration. In those families, cascade testing can provide surveillance and intervention strategies that can be lifesaving.
A related and particularly relevant question is when should a result found on a somatic testing panel prompt follow-up germline testing? Some institutions have algorithms in place to automate referral for germline testing based on specific genetic criteria.12 Excellent reviews are available that outline the following considerations in more detail.13 Typically, somatic testing results that would trigger follow-up germline testing would be truncating or deleterious or likely deleterious mutations per germline datasets in high-risk genes associated with highly penetrant autosomal dominant conditions (BRCA1, BRCA2, PALB2, MLH1, MSH2, and MSH6), selected moderate-risk genes (BRIP1, RAD51C, and RAD51D), and specific variants with a high probability of being germline because they are common germline founder mutations. Although the actionability and significance of specific genes remains a matter of some discussion, generally finding a somatic pathogenic sequencing result included in the 59-gene list of the American College of Medical Genetics and Genomics (ACMG) guidelines would be an indication for germline testing. Another indication for germline testing would be finding genes with germline mutations for which the NCCN has specific management guidelines, or the presence of alterations consistent with known founder mutations.14 When a patient’s tumor has microsatellite instability or is hypermutated (defined as > 10 mutations per megabase), a search for germline alterations is warranted given that about 15% of these patients with these tumors carry a Lynch syndrome gene.15 Genes that are commonly found as somatic alterations alone (eg, TP53 or APC) are generally not an indication for germline testing unless family history is compelling.
Although some clinicians use the variant allele fraction in the somatic sequencing report to decide whether to conduct germline testing, this approach is suboptimal, as allele fraction may be confounded by assay conditions and a high allele fraction may be found in pure tumors with loss of heterozygosity (LOH) of the other allele. There is also evidence that for a variety of reasons, somatic sequencing panels do not always detect germline alterations in somatic tissues.16 Reasons for this may include discordance between the genes being tested in the germline vs the somatic panel, technical differences such as interference of formalin-fixed paraffin-embedded (FFPE) artifact with detecting the germline variant, lack of expertise in germline variant interpretation among laboratories doing tumor-only sequencing, and, in rare cases, large deletions in tumor tissue masking a germline point mutation.