Overview
This lecture covers how genetics is used clinically in cancer, moving from risk to treatment. It starts with the genetic architecture of cancer risk (somatic versus germline change, sporadic versus familial versus hereditary disease), then works through three high-penetrance syndromes in detail (BRCA1/BRCA2 hereditary breast-ovarian cancer, hereditary diffuse gastric cancer and Lynch syndrome), showing for each the gene, the protein function lost, the resulting cancer risk and the surveillance or surgical management that follows. It then recaps breast cancer pathology (histological type, grade, ER/PR/HER2 status) and closes with genomic management of established tumours: targeting HER2 with Herceptin, predicting chemotherapy benefit with Oncotype DX, and tumour panel sequencing to find actionable alterations. The unifying idea is that a defined molecular defect in a tumour or germline predicts both risk and the therapy that will work.
Course context and assessment
Where this sits in Genetics Semester 2 (Assoc. Prof. Anita Dunbier, convenor): it follows the Clinical Genetics lecture (Prof Stephen Robertson, 15 July), the Genetics intro/assignment/HHC lecture and Tutorial 4 (haemochromatosis), the Cancer Genetics lectures (Prof Parry Guilford, 13 and 21 August), the molecular diagnosis and prognosis lecture (CF and cancer, 14 August), Tutorial 5 (CF screening, diagnosis and CF case), and the colorectal clinical demonstration (Prof Chris Jackson, 21 August). It is followed by the Respiratory module (25 August), Tutorial 6 (Cancer Genetics), the Specific Genetic Disorder Report (due 15 Sept, 7pm, 5%) and Tutorial 7 (presentations). Stated linkages: Screening Public Health, CF case, Ethics in Genetics, Blood/iron overload, Clinical Genetics lecture, Cancer vertical module, colorectal cancer case, BIOC192 haemochromatosis lab.
Overarching Genetics aims most relevant to this lecture (aims 4 to 7 of 7): describe the importance of genetic predisposition in relation to multifactorial aetiology of disease; describe approaches to investigation and management of genetic disorders including ethical considerations; describe the somatic and inherited genetic contribution to the pathogenesis of cancer; understand how novel genetic research is interpreted, evaluated in a clinical context and applied to patient care. The other three aims (chromosomal abnormalities, Mendelian disorders, population genetics) belong to earlier lectures.
Aims of Tutorial 6 (Cancer Genomics and the Management of Cancer), which this lecture feeds: carry out a basic assessment of inherited cancer risk; outline the characteristics of inherited cancer and key principles of its clinical management; understand how tumour genome copy number changes and rearrangements can create dependence on specific growth pathways; describe how alterations in the tumour genome produce targets for therapy; understand how gene expression analyses give prognostic information; be aware that primary and metastatic cancers can differ molecularly; describe how analysis of circulating tumour DNA has the potential to monitor metastatic cancer and identify therapies.
The tutorial group talk (week starting 15 Sept) should focus on presentation of the disorder, the red flags that might indicate it, and current and future diagnosis and treatment, pitched so classmates can understand it. It should identify features of interest to the class: likely presentation, interesting genetic aspects, ethical aspects.
Cancer by numbers and the heritability of cancer
Key points:
- Cancer is New Zealand’s biggest killer. Approximately 1 in 3 people will die from cancer.
- Only about 3 to 6 mutations are sufficient to cause initiation and progression, yet some tumours carry thousands of mutations.
- Breast cancer is currently the most common cancer, with a 1 in 8 lifetime risk for women. Colorectal and prostate cancers are also frequent.
- Heritability split: approximately 80% of cancers are sporadic, 5% arise in highly penetrant inherited cancer syndromes, and 15% in lowly penetrant inherited syndromes.
- Penetrance is the likelihood that carriers of a mutation will develop cancer.
Cancer is a genetic disease in two distinct ways:
- Cancer cells contain a large number of alterations from the normal genome, that is, somatic mutations.
- Cancer can cluster within families, through germline pathogenic variants.
Germline versus somatic variants:
| Somatic DNA changes | Germline DNA changes | |
|---|---|---|
| When/where they arise | Acquired over a person’s lifetime, in single cells | Present in every cell of the body, including egg and sperm |
| Effect | Can lead to cancer | Can increase cancer susceptibility |
| Inheritance | Cannot be inherited | Can be inherited |
Genetic architecture of cancer risk: plotting allele frequency against relative risk (1, 2, 5, greater than or equal to 10) gives an inverse relationship. Common variants sit at high allele frequency with low penetrance and low relative risk; rare variants of moderate penetrance sit in the middle; rare variants of high penetrance sit at low allele frequency with high relative risk, forming a long tail.
Distinguishing sporadic, familial and hereditary cancer
Three types of familial association are distinguished: sporadic, familial (family history) and hereditary (family history).
Worked pedigree cases from the lecture:
- Case 1 (sporadic): a single affected woman, breast cancer at 67, with no other affected relatives. One isolated case, no clustering.
- Case 2 (familial): a grandmother with breast cancer at 73 (deceased) and her granddaughter with breast cancer at 67. Two affected relatives on the same side of the family, but no tight hereditary pattern.
- Case 3 (hereditary): four affected individuals on one side of the family across three generations, comprising a grandmother with ovarian cancer at 45 (deceased), her daughter with breast cancer at 52, and two granddaughters, one with breast cancer at 38 and one with two primary breast cancers at 32 and 43. Multiple affected members, early onset, across generations, with multiple primaries in one woman.
Hereditary cancer key points:
- High incidence in family groups.
- Predominance of particular cancer types.
- Early age of onset.
- Increased risk of multifocal or bilateral disease.
- Inherited as autosomal dominant with incomplete penetrance.
The familial pattern is defined by contrast: late age at onset, more affected family members, but no specific pattern of inheritance. The higher risk there is based on environmental and genetic risk factors, not on one specific defective gene passed down through the family.
Classic pedigrees for comparison:
- Classic BRCA1 pedigree: clustering of early-onset breast and ovarian cancer across generations (ovarian dx 49, breast dx 42, ovarian dx 53, breast dx 38).
- Classic BRCA2 pedigree: early onset with a wider spread of cancer types, including male prostate and pancreatic cancer (prostate dx 55, breast dx 45 with pancreatic dx 55, ovarian dx 58, breast dx 52).
High penetrance inherited cancer syndromes
Syndrome, gene(s) and major tissue affected. The three in bold are this lecture’s focus.
| Syndrome | Gene(s) | Major tissue |
|---|---|---|
| Ataxia telangiectasia | ATM | Haemopoietic |
| Bloom syndrome | BLM | Haemopoietic |
| Cowden syndrome | PTEN | Breast, thyroid, endometrial |
| Familial adenomatous polyposis | APC, MUTYH | Colorectal |
| Familial malignant melanoma | CDKN2A, CMM1, CDK4 | Skin |
| Fanconi anaemia | FANC A to L | Haemopoietic, liver |
| Hereditary breast-ovarian cancer | BRCA1, BRCA2 | Breast, ovary |
| Hereditary diffuse gastric cancer (HDGC) | CDH1 | Stomach, breast |
| Lynch syndrome | MLH1, MSH2, MSH6, PMS2 | Colorectal |
| Hereditary papillary renal cell carcinoma | MET | Kidney |
| Juvenile polyposis | SMAD4, BMPR1A, ENG | Gastrointestinal tract |
| Li-Fraumeni | TP53 | Sarcomas, breast, brain |
| Multiple endocrine neoplasia type 1 | MEN1 | Parathyroid, pituitary |
| Multiple endocrine neoplasia type 2 | RET | Thyroid |
| Neurofibromatosis type 1 | NF1 | Neural, optic |
| Neurofibromatosis (NF2 row, see warning) | NF2 | Brain, spinal cord |
| Nevoid basal cell carcinoma | PTC | Skin |
| Nijmegen breakage syndrome | NBS1 | B cells |
| Peutz-Jeghers syndrome | STK11 | Colorectal |
| Retinoblastoma | RB1 | Retina |
| Testicular cancer | TCG1 | Testes |
| Tuberous sclerosis | TSC1, TSC2 | Broad, angiofibromas |
| Von Hippel-Lindau syndrome | VHL | Kidney |
| Xeroderma pigmentosum | XPA, XPC, XPE, ERCC3-5 | Kidney |
Warning
The slide lists two rows as “neurofibromatosis type 1”, the second against gene NF2. This appears to be a labelling error on the original slide (NF2 is the gene of neurofibromatosis type 2); the row is recorded as shown rather than corrected.
Familial breast cancer: BRCA1 and BRCA2
High penetrance genes, both DNA repair genes:
| BRCA1 | BRCA2 | |
|---|---|---|
| Function | DNA repair gene | DNA repair gene |
| Incidence | 1 in 500 to 1000 (1 in 100 in Ashkenazi Jews, 2 founder mutations) | 1 in 600 to 800; also more common in Ashkenazi Jews |
| Breast cancer lifetime risk | 60 to 85% | 40 to 85% |
| Ovarian cancer risk | 40 to 60% | Less than 30% |
Other breast cancer risk genes:
- TP53 (Li-Fraumeni syndrome).
- CDH1 (hereditary diffuse gastric cancer).
- ATM, CHEK2, BRIP1, PALB2, double-stranded DNA damage repair genes, which roughly double breast cancer risk.
- Tens of low penetrance genes.
How loss of BRCA leads to cancer
In a carrier, normal tissue is heterozygous (BRCA1/2 +/mut) while tumour tissue is BRCA1/2 mut/mut. Common tumour sites are breast and ovary; less common sites are stomach, prostate, pancreas, colon and leukaemia.
Wild-type BRCA1/BRCA2 supports three repair functions, all converging on cell survival and maintained genome integrity:
- Homologous recombination repair, with BRCA1, BRCA2 and RAD51 assembling at a DNA break.
- Protection of stalled replication forks, with BRCA1, RAD51 and BRCA2 blocking nuclease degradation of the fork.
- Crosslink repair, with BRCA1, the Fanconi complex and BRCA2 excising the crosslink.
In BRCA1/BRCA2 mutant cells the sequence is:
- Replication fork collapse and DNA breaks.
- Use of non-conservative DNA repair pathways instead (for example single-strand annealing, and NHEJ/alternative end joining).
- Genomic aberrations: radial chromosomes, deletions, translocations.
- Genomic instability, cell death, cancer.
Synthetic lethality and PARP inhibition
BRCA status can be exploited therapeutically. Two repair routes are relevant: base excision repair via PARP1, and homologous recombination via BRCA. Outcomes at a DNA double-strand break:
| Cell state | Base excision repair (PARP1) | Homologous recombination (BRCA) | Outcome |
|---|---|---|---|
| Normal cells | Active | Active | Repair |
| BRCA mutation alone | Active | Inactive | Repair, via the remaining PARP1 pathway |
| Drug-induced PARP1 inhibition alone | Inactive | Active | Repair, via the remaining BRCA pathway |
| BRCA mutation plus PARP1 inhibition | Inactive | Inactive | No repair, cell death |
Losing either pathway alone is survivable; losing both at once is lethal. This is synthetic lethality, and it is the principle exploited by PARP inhibitors in BRCA-mutant tumours.
BRCA1/2 clinical management
- Surveillance: clinical breast exam every 6 to 12 months from 25 years; annual mammography or MRI from about 30 years.
- Prophylactic/risk-reducing surgery: mastectomy gives 90% risk reduction (not complete, because of residual tissue); oophorectomy gives 90% risk reduction if performed before 45 years.
- Chemoprevention: tamoxifen and aromatase inhibitors give 40% breast cancer risk reduction; tamoxifen carries an increased endometrial cancer risk.
Referral criteria for breast cancer genetic testing
From the EVIQ protocols followed by Genetic Services NZ (flagged as relevant to Tutorial 6).
Personal history:
- Triple negative breast cancer (ER, PR and HER2 negative) diagnosed under 50.
- High grade, non-mucinous, epithelial ovarian, fallopian tube or primary peritoneal cancer.
- Lobular breast cancer AND a family history of lobular breast or diffuse-type stomach cancer.
- Breast cancer diagnosed under 40.
- Male breast cancer diagnosed at any age.
- Breast and ovarian primary cancers at any age.
- Two primary breast cancers where the first occurred under 50.
- Breast cancer plus a personal or family history of Peutz-Jeghers syndrome, PTEN hamartoma syndrome or Li-Fraumeni syndrome.
Family history: two first or second degree relatives diagnosed with breast or ovarian cancer, PLUS one or more of the following on the same side of the family:
- Additional relative(s) with breast or ovarian cancer.
- Breast cancer diagnosed under 50.
- More than one primary breast cancer in the same woman.
- Breast and ovarian cancer in the same woman.
- Ashkenazi (Eastern European) Jewish ancestry.
- Male breast cancer.
Hereditary diffuse gastric cancer (HDGC)
Key points:
- Dominant inheritance.
- Incidence 5 to 10 per 100,000, with a 5 times higher rate in Māori.
- Caused by germline CDH1 mutations. CDH1 encodes the cell-to-cell adhesion protein E-cadherin.
- Very rare CTNNA1 mutations also cause it.
- Up to 70% penetrance for advanced diffuse gastric cancer.
- Approximately 40% risk of lobular breast cancer.
The affected structure is the adherens junction between adjacent cell membranes: E-cadherin spans the intercellular gap and is anchored intracellularly by CTNNA1 (alpha-catenin) and beta-catenin, forming repeating junctional units down the membrane.
HDGC clinical management
- CDH1 genetic testing from over 16 years.
- Annual surveillance endoscopy, or prophylactic gastrectomy from over 20 years.
- Annual breast surveillance by MRI.
- Consider risk-reducing mastectomy.
- Different paths depending on family history and severity of mutation.
2020 Wanaka IGCLC genetic testing criteria
Abbreviations: GC gastric cancer; DGC diffuse gastric cancer; HDGC hereditary diffuse gastric cancer; LBC lobular breast cancer; HLBC hereditary lobular breast cancer; PTG prophylactic total gastrectomy; TG total gastrectomy; BRRM bilateral risk-reducing mastectomy.
Pathway: criteria met, then genetic testing for CDH1 and CTNNA1 (alternative testing routes noted for cleft lip/palate, multigene panel test, validated direct-to-consumer test). Three result branches:
- Negative. Ask: does the family history meet genetic testing criteria 1 or 2? If no, likely sporadic DGC or LBC. If yes, classify as HDGC-like and consider annual gastric surveillance for at least 2 years (interval may increase after 2 years; PTG not advised), annual breast surveillance, and breast management based on individualised assessment.
- Positive, that is a pathogenic CDH1 variant carrier. Ask: DGC in the family?
- Yes: HDGC. Recommend PTG; if declined or delayed, offer annual surveillance with reduced emphasis on PTG where the family history is weak, and TG on a positive biopsy. Annual breast surveillance (MRI). Consider bilateral risk-reducing mastectomy with or without reconstruction.
- No: then ask, breast cancer in the family? If yes (with changes to family history), classify as HLBC: annual gastric surveillance, TG on positive biopsy, consider PTG, annual breast surveillance, consider BRRM. If no: annual gastric surveillance and TG on positive biopsy.
- Uncertain, that is a CDH1 variant of unknown significance. Consider annual gastric surveillance for at least 2 years, after which the interval may increase; PTG not advised; breast management based on individualised assessment.
Lynch syndrome (HNPCC)
Key points:
- Lynch syndrome, also called hereditary nonpolyposis colorectal cancer (HNPCC), is an inherited disorder increasing the risk of colon cancer and other cancers including endometrial, ovarian and stomach.
- Prevalence approximately 1 in 300 individuals.
- Key feature: early onset colorectal cancer, often before age 50.
- Caused by inherited mutations in one of the mismatch repair (MMR) genes: MLH1, MSH2, MSH6, PMS2 or EPCAM.
- These genes repair DNA replication errors; mutations lead to microsatellite instability (MSI), increasing cancer risk.
- Autosomal dominant: only one mutated copy is needed to increase cancer risk.
- Tumours show increased responsiveness to immunotherapies.
DNA mismatch repair, in order:
- Mismatch recognition: a mismatched base pair on the daughter/parent duplex is recognised and MutS binds.
- Strand discrimination: MutS is joined by MutL and MutH, identifying which strand is newly synthesised.
- Strand excision and re-synthesis restores the correct sequence.
Microsatellite instability
- Microsatellites are short repetitive DNA sequences (for example , , ) scattered throughout the genome.
- During replication these repetitive sequences are prone to errors such as insertions or deletions, caused by strand slippage.
- In a normal cell the MMR system corrects these errors.
- The resulting instability in microsatellite length is called microsatellite instability (MSI). Unrepaired errors produce either expansion or shortening of the microsatellite.
- Consequently MSI is a surrogate marker for response to immunotherapies.
Neo-antigens
- Neo means new; an antigen is a structure specifically bound by antibodies. Antigens are the unique molecules or proteins that help immune cells identify and fight cancer cells, and neoantigens are unique to each patient’s tumour cells.
- Neoantigens derive from somatic mutations that produce modified or novel peptide sequences within a tumour cell’s repertoire of expressed proteins.
- T-cell receptors (TCRs) are molecules on the surface of cancer-fighting T cells that can interrogate individual cancer cells and see beneath the cell membrane. A T cell binds the neoantigen displayed on the tumour cell surface via its TCR.
Sources of potential neoantigens: single nucleotide variants (SNV), insertions/deletions, spliced peptides, translocations, post-translational modifications. Each alters a DNA/protein sequence which is then processed and presented as a peptide on an MHC-like anchor.
Two experimental approaches identify them:
- Whole exome plus transcriptome sequencing: identification of tumour-specific genomic alterations; determination of clonal mutations; epitope prediction.
- MS-based immunopeptidomics: identification of abundantly expressed mutated ligands; identification of post-translationally modified peptides.
Both converge on selection of potential neoepitopes for therapeutic purposes, leading to: immunogenicity assessment (functional analysis of patient-derived T cells); enrichment of neoantigen-specific T cells; TCR isolation and characterisation (from patients and HLA-matched healthy donors); and vaccination approaches (RNA, peptides, dendritic cells).
Warning
The immune checkpoint blockade slide carries only an unlabelled molecular render of a receptor-ligand interaction bridging two cell membranes at an immune synapse. No specific checkpoint molecules (for example PD-1/PD-L1 or CTLA-4) are named on the slide, so no mechanism detail is available. [slide does not elaborate]
Breast cancer pathology recap
Adenocarcinoma subtypes: ductal carcinoma, lobular carcinoma, tubular carcinoma, medullary carcinoma, mucinous carcinoma, cribriform carcinoma.
Tumour grading and survival:
| Grade | 5 year survival | 7 year survival | Example |
|---|---|---|---|
| Grade 1 | 95% | 90% | Tubular carcinoma |
| Grade 2 | 75% | 63% | |
| Grade 3 | 50% | 45% | Ductal carcinoma |
Receptor status:
- In modern practice approximately 75% of breast cancers are oestrogen receptor positive (ER+).
- Oestrogen antagonism or withdrawal reduces tumour cell proliferation in ER+ breast cancers.
- The progesterone receptor is also prognostic for response to endocrine therapy.
- HER2 overexpression is both prognostic and predictive. Approximately 25% of breast cancers are HER2 positive.
Oestrogen signalling
Steps in order:
- Oestradiol crosses the cell membrane.
- It binds the oestrogen receptor.
- Receptor-hormone complexes dimerise.
- The dimer translocates to the nucleus, where it takes one of two routes:
- Recruit a corepressor complex (Sin3, NCOR, HDACs), silencing transcription.
- Bind the oestrogen response element (ERE) with coactivators (TAFs, TBP, p300, CBP, TFII, RNA Pol II), driving transcription of Myc, Cyclin D1, Cyclin E1 and Cyclin E2.
- The transcriptional route leads to cell growth and proliferation.
HER2 as a target
- The normal breast cell carries sparse HER2 receptors, transcribed from a single-copy ERBB2 gene on chromosome 17; growth factor binding triggers activation of signalling pathways into the nucleus.
- In HER2 positive cancer the ERBB2 gene segment on chromosome 17 is amplified, producing much more messenger RNA and a cell surface densely studded with HER2 protein, so growth factor binding drives excessive growth signalling.
- Herceptin, an antibody, binds the HER2/neu receptors: it prevents growth factor binding and activates an immune response, giving no signalling into the nucleus, so growth slows.
Interpreting pathology reports
- Different scoring systems may be used, for example the Allred score or percentage staining.
- It is crucial to know whether a result is positive, negative or other.
- Staining may be reported as strong or weak.
- HER2 is evaluated by immunohistochemistry as 1+, 2+ or 3+.
- If 2+ (equivocal), FISH (fluorescence in situ hybridisation) is needed to determine whether gene amplification has occurred. The level of the gene of interest is compared to a control probe to give a relative amount.
- On ERBB2 FISH, a positive nucleus shows dense clusters of red signal, many more red than green; a negative nucleus shows sparse, roughly equal red and green signals, that is a normal red:green ratio.
Genomic management of cancer: Oncotype DX
Background:
- Paik et al. (2004), NEJM 351, 2817.
- Developed commercially from a candidate panel of 250 genes.
- Current list price USD 4,620.
- Since 2004 used by over 10,000 physicians to guide treatment for over 200,000 patients in more than 60 countries.
- It answers the clinical question “chemo or no chemo?”.
The assay:
- PCR-based (RT-PCR) quantification of tumour mRNA for 21 genes, producing a Recurrence Score between 0 and 100.
- Six gene categories feed the score: Proliferation (Ki-67, STK15, Survivin, Cyclin B1, MYBL2); Invasion (Stromelysin 3, Cathepsin L2); HER2 (GRB7, HER2); Estrogen (ER, PR, Bcl2, SCUBE2); Reference (Beta-actin, GAPDH, RPLPO, GUS, TFRC); Other (GSTM1, CD68, BAG1).
Basic risk bands: less than 18 low recurrence risk; 18 to 25 moderate; greater than 25 high.
Uses:
- Assess risk of recurrence in early-stage ER+ breast cancer, and the likely benefit from chemotherapy after breast cancer surgery.
- Assess risk of DCIS (ductal carcinoma in situ) recurrence and/or risk of a new invasive cancer developing in the same breast, plus likelihood of benefit from radiation therapy after DCIS surgery.
Age-stratified chemotherapy decision thresholds:
| Age group | Recurrence Score | Interpretation |
|---|---|---|
| Older than 50 | 0 to 25 | Low risk of recurrence; chemotherapy benefits likely will not outweigh side-effect risks |
| Older than 50 | 26 to 100 | High risk; chemotherapy benefits likely greater than the risks |
| 50 and younger | 0 to 15 | Low risk; chemotherapy benefits likely will not outweigh the risks |
| 50 and younger | 16 to 20 | Low to medium risk; chemotherapy benefits likely will not outweigh the risks |
| 50 and younger | 21 to 25 | Medium risk; chemotherapy benefits likely greater than the risks |
| 50 and younger | 26 to 100 | High risk; chemotherapy benefits likely greater than the risks |
Validation data (Dowsett et al., J Clin Oncol 2010; 28(11):1829-1834)
Proportion distant recurrence free over 0 to 9 years, split by Low/Intermediate/High score group, log-rank P less than .001 in both cohorts.
| Cohort | Group | n (%) | Events | Approx. endpoint recurrence-free | Hazard ratio (95% CI) |
|---|---|---|---|---|---|
| Node negative | Low | 513 (59%) | 20 | ~96% | reference |
| Node negative | Intermediate | 229 (26%) | 24 | ~88% | 2.5 (1.3 to 4.5) v Low |
| Node negative | High | 130 (15%) | 28 | ~75% | 5.2 (2.7 to 10.1) v Low |
| Node positive | Low | 160 (52%) | 25 | ~83% | reference |
| Node positive | Intermediate | 94 (31%) | 25 | ~72% | 1.8 (1.0 to 3.2) v Low |
| Node positive | High | 52 (17%) | 24 | ~51% | 2.7 (1.5 to 5.1) v Low |
Example clinical report
A sample Oncotype DX Breast Recurrence Score report shows a Recurrence Score result of 17. The report states that the test uses RT-PCR to determine expression of a 21-gene panel in tumour tissue, with results ranging 0 to 100, and that the findings apply to women with ER+ breast cancer and 1 to 3 positive nodes who will be treated with 5 years of hormonal therapy. The clinical experience data come from the SWOG 8814 validation study of 367 post-menopausal patients with node-positive, ER+ breast cancer randomised to tamoxifen alone or CAF chemotherapy followed by tamoxifen, with 5-year disease-free survival as the endpoint. Three studies in more than 4,000 contemporary patients with 1 to 3 positive nodes reported low rates (under 4%) of 5-year distant recurrence and/or breast-cancer-specific mortality in patients with Recurrence Score results below 12 (PlanB) and below 18 (Clalit and SEER) treated with hormonal therapy alone (tamoxifen or an aromatase inhibitor). For 1 to 3 positive nodes, the 5-year risk of recurrence or mortality was 11% (95% CI 7% to 18%) with tamoxifen alone and 12% (95% CI 8% to 18%) with tamoxifen plus chemotherapy.
Health-system impact (BioWorld, 26 April 2019, Annette Boyle): a study by the Georgetown Lombardi Comprehensive Cancer Center and the National Cancer Institute, published in JNCI, indicated that Oncotype DX-guided treatment could cut the cost of first-year breast cancer care in the US by USD 50 million and allow 70% of breast cancer patients to skip chemotherapy. Oncotype DX is produced by Genomic Health Inc. of Redwood City, California.
Cancer precision medicine via panel gene sequencing
Panel sequencing tests many cancer genes at once in a tumour sample. FoundationOne sequences 315 cancer genes per tumour, plus a separate set of genes tested for structural rearrangements. Genes on the panel relevant to this lecture include BRCA1, BRCA2, ATM, CHEK2, PALB2, MLH1, MSH2, MSH6, PMS2, PTEN, TP53, RB1, VHL, NF1, NF2, MET, EGFR and ERBB2 (HER2). [The full panel gene list on the slide is not exhaustively transcribed; it contains well over a hundred further gene symbols in small print.]
What a report contains, using the lecture’s sample FoundationOne report (a bladder urothelial/transitional cell carcinoma):
- Header fields: patient name, date of birth, sex, report date, medical facility, ordering physician, tumour type, specimen received/site/date of collection/type, FMI case number, medical record number, medical facility ID, specimen ID.
- About the test: FoundationOne is a next-generation sequencing (NGS) based assay identifying genomic alterations within hundreds of cancer-related genes.
- Patient results summary: 9 genomic findings; 9 therapies associated with potential clinical benefit; 0 therapies associated with lack of response; 10 clinical trials.
- Genomic alterations identified: EGFR amplification; LRP1B splice site 1553-2A>T and splice site 2887+1G>T; NOTCH2 H1300fs15; RBM10 K653fs51; SETD2 splice site 6061-1G>T; SMARCA4 Q338*; TP53 R337L.
- Additional findings: tumour mutation burden TMB-High, 20.76 mutations/Mb.
- Therapeutic implications table, listing for each finding the FDA-approved therapies in the patient’s own tumour type, FDA-approved therapies in another tumour type, and potential clinical trials:
- TMB-High (20.76 Muts/Mb): none in this tumour type; atezolizumab, nivolumab, pembrolizumab in another tumour type; clinical trials available.
- EGFR amplification: none in this tumour type; afatinib, cetuximab, erlotinib, gefitinib, lapatinib, panitumumab in another tumour type; clinical trials available.
- LRP1B splice site variants: none, none, none.
- NOTCH2 H1300fs*15: none, none, none.
Important
The pattern to take from the sample report is that most actionable findings led to therapies approved in a different tumour type rather than the patient’s own, so the practical output of panel sequencing is often an off-label option or a clinical trial rather than a licensed treatment.
Self-test
- State the approximate proportions of cancers that are sporadic, highly penetrant inherited and lowly penetrant inherited, and define penetrance.
- Explain the two distinct senses in which cancer is described as a genetic disease.
- Distinguish somatic from germline DNA changes on when they arise, their effect and their heritability.
- Describe the relationship between allele frequency and relative risk in the genetic architecture of cancer risk.
- List the five key features of hereditary cancer.
- Given a pedigree with a grandmother diagnosed with breast cancer at 73 and a granddaughter at 67 and no other affected relatives, classify the pattern and justify the classification.
- Name the gene(s) and major tissue for Lynch syndrome, hereditary diffuse gastric cancer and hereditary breast-ovarian cancer.
- Compare BRCA1 and BRCA2 on population incidence, breast cancer lifetime risk and ovarian cancer risk.
- List the three DNA repair functions of wild-type BRCA1/BRCA2.
- Describe, in order, what happens in a BRCA-mutant cell once its repair functions are lost.
- Explain why a PARP1 inhibitor kills BRCA-mutant tumour cells but spares normal cells.
- State the risk reduction achieved by prophylactic mastectomy, by oophorectomy and by chemoprevention in BRCA1/2 carriers, including any condition attached to each figure.
- A 38-year-old woman is diagnosed with triple negative breast cancer. Explain whether she meets referral criteria for breast cancer genetic testing on personal history alone, and why.
- Describe the molecular defect in hereditary diffuse gastric cancer, including the gene, the protein and the structure affected.
- State the penetrance for advanced diffuse gastric cancer and the lobular breast cancer risk in HDGC, and give the management offered for each of these two risks.
- A patient carries a pathogenic CDH1 variant and there is diffuse gastric cancer in the family. Outline the recommended management under the 2020 Wanaka IGCLC pathway.
- Distinguish the management of a pathogenic CDH1 variant carrier from that of a carrier of a CDH1 variant of unknown significance.
- Name the mismatch repair genes involved in Lynch syndrome and describe the three steps of mismatch repair.
- Explain how loss of mismatch repair produces microsatellite instability, and why MSI predicts response to immunotherapy.
- Define a neoantigen and list the sources from which neoantigens can arise.
- Describe how Herceptin works, and state the underlying genomic abnormality it targets.
- A breast tumour scores HER2 2+ on immunohistochemistry. Describe the next test and how its result is interpreted.
- Give the Oncotype DX chemotherapy decision thresholds for a woman aged 55 and for a woman aged 45.
- Explain what the Dowsett validation data show about the prognostic value of the Recurrence Score in node-negative versus node-positive disease.
- Integrative: for BRCA1/2 mutation, Lynch syndrome and HER2 amplification, state in each case the molecular defect and the specific therapeutic strategy it makes possible.
Answers
Reveal answers
- Approximately 80% of cancers are sporadic, 5% arise in highly penetrant inherited cancer syndromes and 15% in lowly penetrant inherited syndromes. Penetrance is the likelihood that carriers of a mutation will develop cancer.
- First, cancer cells contain a large number of alterations from the normal genome, that is somatic mutations. Second, cancer can cluster within families through germline pathogenic variants.
- Somatic changes are acquired over a lifetime in single cells, can lead to cancer, and cannot be inherited. Germline changes are present in every cell including egg and sperm, can increase cancer susceptibility, and can be inherited.
- They are inversely related. Common variants have high allele frequency but low penetrance and low relative risk; rare variants of moderate penetrance sit in the middle; rare variants of high penetrance have low allele frequency and high relative risk, forming a long tail.
- High incidence in family groups; predominance of particular cancer types; early age of onset; increased risk of multifocal or bilateral disease; inherited as autosomal dominant with incomplete penetrance.
- Familial. There are two affected relatives on the same side of the family, but with late age at onset and no specific pattern of inheritance; the raised risk reflects environmental and genetic risk factors rather than one specific defective gene passed down the family.
- Lynch syndrome: MLH1, MSH2, MSH6, PMS2, colorectal. Hereditary diffuse gastric cancer: CDH1, stomach and breast. Hereditary breast-ovarian cancer: BRCA1 and BRCA2, breast and ovary.
- BRCA1 incidence 1 in 500 to 1000 (1 in 100 in Ashkenazi Jews, with 2 founder mutations), breast cancer lifetime risk 60 to 85%, ovarian cancer risk 40 to 60%. BRCA2 incidence 1 in 600 to 800 and also more common in Ashkenazi Jews, breast cancer lifetime risk 40 to 85%, ovarian cancer risk under 30%.
- Homologous recombination repair (with RAD51 at a DNA break); protection of stalled replication forks from nuclease degradation; and crosslink repair (with the Fanconi complex). All three maintain cell survival and genome integrity.
- Replication fork collapse and DNA breaks; use of non-conservative repair pathways instead, such as single-strand annealing and NHEJ/alternative end joining; genomic aberrations including radial chromosomes, deletions and translocations; then genomic instability, cell death and cancer.
- Double-strand break repair can proceed by base excision repair via PARP1 or by homologous recombination via BRCA. Normal cells have both, so blocking PARP1 alone still allows repair via BRCA. BRCA-mutant cells have already lost homologous recombination, so adding PARP1 inhibition removes the only remaining route, giving no repair and cell death. This dual loss is synthetic lethality.
- Prophylactic mastectomy gives 90% risk reduction, not 100%, because of residual breast tissue. Oophorectomy gives 90% risk reduction if performed before 45 years. Chemoprevention with tamoxifen or aromatase inhibitors gives 40% breast cancer risk reduction, with tamoxifen carrying an increased endometrial cancer risk.
- Yes. Triple negative breast cancer (ER, PR and HER2 negative) diagnosed under the age of 50 is itself a personal-history referral criterion, and breast cancer diagnosed under 40 is a separate criterion she also meets.
- Germline mutation in CDH1, which encodes E-cadherin, the cell-to-cell adhesion protein. E-cadherin spans the intercellular gap at the adherens junction and is anchored intracellularly by CTNNA1 (alpha-catenin) and beta-catenin. Very rare CTNNA1 mutations cause the same syndrome.
- Up to 70% penetrance for advanced diffuse gastric cancer and approximately 40% risk of lobular breast cancer. Gastric risk is managed by annual surveillance endoscopy or prophylactic gastrectomy from over 20 years; breast risk by annual breast surveillance with MRI and consideration of risk-reducing mastectomy.
- This is HDGC. Recommend prophylactic total gastrectomy; if it is declined or delayed, offer annual surveillance with reduced emphasis on PTG where the family history is weak, and total gastrectomy on a positive biopsy. Also annual breast surveillance by MRI and consideration of bilateral risk-reducing mastectomy with or without reconstruction.
- A pathogenic variant carrier with DGC in the family is recommended prophylactic total gastrectomy plus annual breast surveillance and possible BRRM. For a variant of unknown significance, PTG is not advised; instead consider annual gastric surveillance for at least 2 years, after which the interval may increase, with breast management based on individualised assessment.
- MLH1, MSH2, MSH6, PMS2 and EPCAM. Steps: mismatch recognition, where MutS binds the mismatched base pair; strand discrimination, where MutS is joined by MutL and MutH; then strand excision and re-synthesis restoring the correct sequence.
- Microsatellites are short repetitive sequences prone to insertion or deletion errors from strand slippage during replication. Normally MMR corrects them; without MMR the errors persist, so repeat tracts expand or shorten, which is microsatellite instability. MSI is a surrogate marker for response to immunotherapies, and Lynch syndrome tumours show increased responsiveness to immunotherapy.
- A neoantigen is a new antigen derived from somatic mutations that produce modified or novel peptide sequences within a tumour cell’s repertoire of expressed proteins, unique to each patient’s tumour. Sources: single nucleotide variants, insertions/deletions, spliced peptides, translocations and post-translational modifications.
- Herceptin is an antibody that binds the HER2/neu receptor, preventing growth factor binding and activating an immune response, so there is no signalling into the nucleus and growth slows. The target abnormality is amplification of the ERBB2 gene on chromosome 17, which produces excess messenger RNA and a cell surface densely studded with HER2 protein.
- A 2+ result is equivocal, so FISH (fluorescence in situ hybridisation) is performed to determine whether gene amplification has occurred. The level of the gene of interest is compared to a control probe to give a relative amount: a positive nucleus shows dense red clusters with many more red than green signals, a negative nucleus shows sparse, roughly equal red and green signals.
- Aged 55 (over 50): score 0 to 25 is low risk and chemotherapy benefits likely will not outweigh the side-effect risks; 26 to 100 is high risk and benefits likely outweigh the risks. Aged 45 (50 and younger): 0 to 15 low and 16 to 20 low to medium, both with benefits unlikely to outweigh the risks; 21 to 25 medium and 26 to 100 high, both with benefits likely greater than the risks.
- The score separates outcomes strongly in both settings, with log-rank P below .001. In node-negative disease the separation is wider: distant-recurrence-free proportions of about 96%, 88% and 75% for Low, Intermediate and High, with hazard ratios of 5.2 (High v Low) and 2.5 (Intermediate v Low). In node-positive disease outcomes are worse overall and the discrimination is smaller: about 83%, 72% and 51%, with hazard ratios of 2.7 and 1.8.
- BRCA1/2 mutation means loss of homologous recombination repair, which allows synthetic lethality with a PARP inhibitor that blocks the remaining base excision repair route. Lynch syndrome means loss of mismatch repair, producing microsatellite instability and a high mutation burden with neoantigens, which makes the tumour responsive to immunotherapy. HER2 amplification means ERBB2 copy gain and receptor overexpression, which allows antibody blockade with Herceptin to prevent growth factor binding and activate an immune response.