VAF and neoplastic cell content explained

In the previous step, we came across the terms ‘VAF’ and ‘neoplastic cell content’. In this step, we will learn what they are and what they can tell us.

Introducing the terms

• Neoplastic cell content is an indication from the pathologist about how many tumour cells are present within the piece of tissue, and it tells us something about the sensitivity of the test.
• Variant allele frequency, or VAF, tells us what proportion of variant DNA has been detected in the sample.

But what are the implications of these pieces of information for the referring clinician?

Neoplastic cell content

The sensitivity of a test is based on the number of tumour cells that are in the sample. Let’s think about this in context with a patient who has a colorectal tumour.

In figure 1 below, the blue and yellow circles represent cells within our tissue sample – the blue cells are tumour cells and the yellow cells are infiltrating normal, or stromal, cells. So, if we were looking for, for example, a KRAS variant in a patient, what would we expect to find?

_{Figure 1: Cells within a colorectal tumour sample, including nine tumour cells (blue) and seven ‘normal’ or stromal cells (yellow), each with two copies of the KRAS gene. Each ‘normal’ cell has two wild-type copies of the KRAS gene, while each tumour cell has one wild-type and one variant copy of the gene.}

The neoplastic cell content in this example is 56%, as 9 out of 16 cells are tumour cells. This impacts on the detection of the KRAS variant.

Every cell has two copies of each chromosome including the chromosome that has the KRAS gene. If we look at our normal cells highlighted in yellow, both copies of the KRAS gene indicated by the white squares will be typical or wild-type, while the tumour cells have one copy of the KRAS gene with the variant and one that is wild-type.

In this example, 28% of copies of the KRAS gene in the sample have the variant. It is present in one copy of all the tumour cells, but not at all in the non-tumour cells. It’s important to have the context given to us by the pathologist around how many tumour cells compared to non-tumour cells there are within the piece of tissue, so that we can understand the sensitivity of the test and interpret the data when sequencing a tumour sample.

Variant allele frequency (VAF)

Many hundreds, and more often thousands, of individual DNA sequences called reads are generated when we sequence a genomic sample. As a result, we end up with a mixture of normal reads and variant reads. The variant allele frequency refers to the proportion of variant reads within the sample. This can be a crucial piece of information for interpreting the report. It’s important to note that the variant allele frequency can be influenced by a number of factors and is not always straightforward to interpret.

We’ve already covered neoplastic cell content, but VAF can also be influenced by whether the variant developed early or late in cancer development, by treatment factors and by the genomic instability of the cancer because of deletions, amplifications or duplications. It can also be impacted by loss of heterozygosity and whether the variant is constitutional (germline) or somatic in origin.

When constitutional (germline) variants are identified

Sometimes, a report can identify constitutional (germline) variants, but how? And what are the implications when this happens? In the case below, we have a patient who has a metastatic castrate-resistant prostate cancer and has been analysed for the usual gene panel according to the test directory, M218.

_{Figure 2: A genomic test report that has identified a constitutional (germline) result – a pathogenic variant in the ATM gene.}

From the result field, we can see that the patient had a variant detected in the ATM gene with a deletion of five nucleotides, which is a frameshift mutation. It’s also important to note the variant allele frequency, which is at 55 percent. This means that over half of the DNA in the sample has the variant in it. This would only be likely if the tumour sample was made up entirely of tumour cells, or it could indicate that this variant may be constitutional and we are seeing the variant in all cells, not just the tumour cells.

In this report, the recommendation from the laboratory is that a blood sample should be taken and analysed to determine if the variant is constitutional in origin. In addition, they recommend that the patient is referred to clinical genetics because if they have a constitutional variant, that means that further discussions might need to take place around their ongoing cancer risk and whether testing might be appropriate for other family members.

In this case, further testing did identify that the ATM variant was constitutional in origin and the patient’s mother was confirmed to have had ovarian cancer at age 40. This fits with there being a family history of the ATM variant, and would mean that other family members could be tested for the same variant.

So, now we’ve now worked through a genomic test report and all of the elements you might expect to see, it’s over to you for a short quiz to consolidate your learning.