Comparing Somatic Cancer Panels

The number of companies offering large panel somatic testing for cancer is growing rapidly, as are the types of panels available. It can sometimes be difficult to determine the important differences. The following list provides some questions and information about test characteristics that may help differentiate among available tests.

For which types of cancer is the panel best?

Labs have different processes to determine which genes to include on the panels they offer. In most cases, genes that are associated with targeted treatments, either FDA-approved or in active clinical trials, are included. There may be other genes included as well, based on other, lab-specific criteria. 

Sensitivity and specificity of the panel vary depending on characteristics of the tumor, including stage, vascularity, and heterogeneity. Some tests are only validated for solid tumors while others are specific for hematologic or cancer types seen most frequently in pediatric cases. In addition, evidence of actionability may be stronger for certain tumor types than others. 

What type of variants does the panel detect? (e.g., SNV, CNV, fusions)

Not all test platforms are able to detect the same type of variants. For example, a platform validated to detect changes of a single or a few nucleotides within a gene cannot detect a fusion between two chromosomes, or changes in RNA or protein expression. Some labs provide testing options that analyze the sample on more than one platform in order to provide information about different types of variants. Knowing what type of variants the test can detect is important to ensure that you will get the information you expect.

Does the test detect all variants within a gene or specific known variants?

There are different approaches to developing somatic cancer testing, each with benefits and limitations. All of these testing types can be accomplished with next generation sequencing (NGS) platforms, so knowing that NGS is the test platform alone is not enough information to distinguish test scope.

• Targeted panels detect a specific set of variants or region of interest. Which variants a specific panel detects depend on its purpose and construction.
  • Hotspot panels typically include specific, individual variants commonly seen in tumors and associated with FDA-approved treatment.
  • Other panels assess entire genes to identify any variants that might be present. The number of genes included in these panels is highly variable, from only a few to hundreds
• Whole exome sequencing assesses the protein coding regions (exons) of all genes in the genome, unselected for those that are associated with cancer.
• Whole genome sequencing assesses all coding and non-coding regions (exons and introns) across the genome.

There are benefits and limitations to each of these tests.

• Hotspot panels often provide results that are immediately applicable to determining treatment, but will not provide additional options if the tumor does not have variants associated with the few treatments approved by the FDA.
• The broader the panel, the greater the chance that variants will be identified, but some variants are likely to have unknown significance or be associated with treatments for which the evidence for efficacy in the patient’s cancer type is less robust.

How likely is it that the test will detect a variant if it is present?

Some oncology providers may consider using a large panel in place of tests that analyze only a few recommended variants, such as EGFR variants in non-small cell lung cancer. This may be an option but the ability of the larger pale to detect certain variants will depend on its sensitivity and specificity..  When targeting specific variants, the ordering provider should confirm that the large panel test is able to detect these variants with the same sensitivity and specificity as the individual tests.

The testing platform affects the sensitivity and specificity of detection. In general, broader tests have somewhat lower sensitivity than narrow, more focused tests. For example, whole genome sequencing will have a lower sensitivity for any one particular variant than a lung cancer panel that tests for only a few specific variants. That said, the sensitivity of larger panels is often sufficient to adequately detect the variants of interest.

What clinically actionable variants does the lab report?

There are several categories of “actionable” variants that a laboratory may report.  These include treatments associated with:

• An FDA-approved therapy for the patient’s cancer type
• An FDA-approved therapy for a different diagnosis
• Therapies in active clinical trials

A recent consensus statement from the Association for Molecular Pathology, American Society of Clinical Oncology, and College of American Pathologists suggests how variants should be categorized based on their clinical impact. 

How does the lab identify clinical trials?

Each lab uses its own algorithms and processes to identify clinical trials. One difference among reporting practices is the level of specificity of the identified trials. Some labs may provide any trial that is testing a treatment related to a gene in which a variant is identified. Other labs may filter trials and provide a smaller list, more targeted to the specific variants identified. There also may be differences in filtering based on the characteristics of the cancer diagnosis (e.g., breast cancer vs. triple negative breast cancer). In all cases, the ordering clinician will need to further investigate the clinical trials listed to determine which, if any, the patient is eligible for based on additional clinical information.

Does the lab report variants of unknown significance?

Genomic variants are abundant in tumor cells; some are benign, while others disrupt or alter the function of the gene and affect tumor growth.  Others have unknown or unclear effects on the cancer cells; these are called variants of uncertain significance, or VUS.

Laboratories vary in their reporting practices for VUS. In some cases, such as hotspot panels, the test does not detect VUS. In panels where the entire gene is assessed, the lab may or may not choose to report VUS. In this case, when labs do not provide the information on the report, they may be able to provide it when asked by the ordering clinician. Although VUS are not actionable, some clinicians want the information in case more data becomes available in the future. 

Does the test include assessing normal as well as tumor tissue?

Genomic variants are present in normal as well as tumor cells. The ultimate goal of somatic testing is to identify the variants that are driving the growth and proliferation of the tumor and associated targeted treatments. Doing this requires filtering out variants that are unlikely to be contributing to cancer cell growth.

Labs differ in whether they use a tumor sample alone or a tumor sample in combination with a normal sample to help interpret results. Tumor-only approaches rely on databases that record population frequency of variants that occur in the germline by chance alone, not related to disease. Existing databases may not represent all ethnicities and may not have information about frequency of all variants. Tumor-normal approaches sequence both tumor and normal DNA to help determine which variants are unique to the tumor. This approach requires sequencing of two full genomes (or parts thereof, depending on approach used), which significantly increases the cost of the test. The best approach is currently an area of active discussion.

Does the lab report germline variants and secondary findings?

If the test includes assessing for variants in normal cells as well as tumor cells, it may identify germline variants associated with hereditary syndromes. Labs vary in whether they report these germline variants or just use them for filtering; most commercial labs currently do not report germline variants. Any germline variant identified needs to be confirmed with additional, validated testing.

Secondary findings are germline variants that are associated with increased risk for conditions unrelated to the test indication. Recently, the American College of Medical Genetics and Genomics developed a list of 59 genes associated with hereditary syndromes that are clinically actionable. They recommend that labs report pathogenic variants found in any of these genes, regardless of the test indication.

It is important to know whether the lab reports germline variants and, specifically, secondary findings because it will impact pre-test counseling and, potentially, the care plan and post-test discussion.

Test Practicalities

Lab practices can differ significantly with regard to the logistics of test ordering and reporting as well as available support. The following questions address some of these differences.

• What are the sample requirements (how much tissue is needed, formalin-fixed paraffin-embedded or fresh frozen, blocks or slides, etc.)?
• How many genes/variants are included on the panel?
• Does the panel include the specific gene/variant you are looking for?
• How well does the panel detect all classes of genetic alterations (base substitution, insertions/deletions, copy number alterations, rearrangements)?
• Do the results include variants of uncertain significance?
• Are copy number and allele frequency provided for alterations?
• Are tumor mutation burden and microsatellite stability reported?
• What support is provided with test selection and interpretation?
• What support is provided with insurance authorization?
• What clinical information is important to include on requisitions?
• What is the cost of testing?
• What is the expected turnaround time?