Tumour Mutation Testing (BRAF, KRAS, EGFR)

Tumour mutation testing finds acquired changes in the DNA of cancer cells that drive growth and predict response to treatment. BRAF, KRAS and EGFR are among the most commonly tested genes, since their mutations decide the choice of targeted treatment in several cancers.

The test is done on DNA extracted from tumour tissue (formalin fixed, paraffin embedded) or cytology material, and sometimes on cell free DNA in blood. Methods are real time PCR hotspot assays, pyrosequencing, Sanger sequencing, and next generation sequencing. Rapid hotspot tests give results in 2 to 5 days, and sequencing panels in 7 to 14 days.

A pathologist selects an area with enough tumour (usually at least 10 to 20% tumour cells, and more for less sensitive methods) and marks it for extraction. The result is reported as the gene, the exact change (for example, BRAF V600E), and its clinical meaning.

Mutations are frequent in some tumours. BRAF V600E is present in about 40 to 50% of melanomas, about 45% of papillary thyroid cancers, 8 to 10% of colorectal cancers, and 1 to 3% of lung adenocarcinomas. KRAS is mutated in about 40% of colorectal cancers, 25 to 30% of lung adenocarcinomas, and over 90% of pancreatic cancers. EGFR is mutated in about 10 to 15% of lung adenocarcinomas in Western populations and 30 to 50% in East Asian populations.

Indications.

• Metastatic colorectal cancer: RAS (KRAS and NRAS) and BRAF testing before any anti-EGFR antibody treatment, and at diagnosis for prognosis.
• Non-small cell lung cancer (adenocarcinoma, and non-squamous or never smoker squamous types): EGFR, KRAS, BRAF and other genes, before treatment.
• Melanoma (advanced or high risk resected): BRAF V600 testing, which decides use of BRAF and MEK inhibitors.
• Papillary and anaplastic thyroid cancer: BRAF V600E for diagnosis, prognosis and treatment.
• Pancreatic, biliary and other cancers: KRAS and other genes, increasingly as part of a panel.
• Hairy cell leukaemia (BRAF V600E), Langerhans cell histiocytosis and Erdheim-Chester disease.
• Progression on targeted treatment: repeat testing for resistance mutations (for example, EGFR T790M, C797S).

Not indicated for early stage disease where results do not change treatment, or when tissue is inadequate and a repeat sample or blood test is more suitable.

Specimen requirements.

• Formalin fixed, paraffin embedded tumour tissue or cell block, fixed in 10% neutral buffered formalin for 6 to 72 hours. Strong acid decalcification is avoided, with EDTA used for bone.
• Tumour cell content assessed by a pathologist, with macrodissection when the content is low.
• Alternatively, plasma cell free DNA collected in special tubes, when tissue is unavailable or unsafe to collect.
• Clinical details, tumour type, stage and treatment history.

Normal appearances.

A negative result ("no mutation detected") means no variant was found in the regions tested, at a level over the detection limit of the method. Normal germline sequence is reported as wild type.

Interpretation.

• BRAF V600E and other V600 mutations: sensitive to BRAF inhibitors given with MEK inhibitors in melanoma, lung cancer and thyroid cancer. In colorectal cancer, V600E carries a poor prognosis, and treatment combines BRAF and EGFR inhibition with or without chemotherapy.
• BRAF non-V600 mutations: different biology, with variable drug responses.
• KRAS and NRAS (codons 12, 13, 59, 61, 117 and 146) in colorectal cancer: predict lack of response to anti-EGFR antibodies. A tumour that is wild type for RAS and BRAF may respond.
• KRAS G12C in lung and colorectal cancer: targetable by specific inhibitors.
• EGFR exon 19 deletions and L858R: sensitise lung adenocarcinoma to EGFR tyrosine kinase inhibitors. Rarer variants (G719X, L861Q, S768I) are also sensitive. Exon 20 insertions need different agents.
• EGFR T790M: appears in resistance to first and second generation inhibitors, and responds to a third generation inhibitor.
• Variants of uncertain significance: reported without a treatment implication, and reviewed by the multidisciplinary team.
• Variant allele frequency: a low frequency suggests a subclonal mutation, a low tumour content, or a germline origin when near 50%.
• Mutations in different genes in the same growth signalling route are usually mutually exclusive, so two driver mutations together suggest a testing issue or an unusual tumour.

Critical values and findings needing immediate action.

• A targetable mutation in a patient with advanced cancer: urgent referral to the oncology team to start the matched treatment.
• RAS mutation in a patient being considered for an anti-EGFR antibody: the antibody is withheld.
• Failure of the test (insufficient tumour or DNA): repeat testing on another block, or liquid biopsy, without delaying treatment decisions.
• Germline like variant frequency in a gene linked to hereditary cancer: referral for genetic counselling.

Limitations and artefacts.

• Tumour content under the assay threshold gives false negatives. Sensitivity differs between methods: Sanger sequencing needs about 15 to 20% mutant allele, while hotspot PCR and sequencing panels detect 1 to 5%.
• Formalin damages DNA (cytosine deamination), producing false low level variants, and prolonged fixation, decalcification and old blocks reduce quality.
• Tumour heterogeneity: a mutation may be present in one area and not another, and primary and metastatic sites can differ.
• Hotspot panels miss rare variants outside the covered regions.
• Interpretation of variants of uncertain significance is difficult, and databases change with time.
• Cell free DNA tests have lower sensitivity (a negative blood test does not exclude a mutation), and are less reliable in tumours that shed little DNA.
• Turnaround time (a delay of 1 to 2 weeks) can hold up treatment, so tests are requested early and in parallel with staging.
• Cost, access, and the need for validated methods limit testing.
• A mutation does not guarantee response, and resistance develops over time.

Special groups.

• Never smokers and East Asian patients with lung adenocarcinoma: higher EGFR mutation rates, and testing in all patients.
• Young patients with colorectal cancer or a family history: mismatch repair and germline testing in addition to mutation testing.
• Patients with small biopsies and cytology samples: cell block tested with limited material, and liquid biopsy used as a back up.
• Patients progressing on targeted treatment: new biopsy or plasma test for resistance mutations.
• Pregnant women: treatment decisions are individual.
• Regions: molecular testing is available in few laboratories, and samples are sent to a reference centre, with delays and cost to patients. Single gene hotspot assays (for example, EGFR and KRAS) give most of the benefit and are cheaper than panels. Targeted drugs are expensive, so access to treatment may limit the use of the result.

Follow up tests.

• Next generation sequencing panels, including ALK, ROS1, RET, NTRK, MET and HER2 in lung cancer.
• Microsatellite instability and mismatch repair testing in colorectal cancer.
• PD-L1 immunohistochemistry.
• Repeat testing in plasma or on a new biopsy at progression.
• Germline genetic testing and counselling when indicated.
• Staging imaging and treatment planning at the multidisciplinary meeting.
• Monitoring of response by imaging, and circulating tumour DNA where available.