Cancer genomics — when your genome turns against you

Driver mutations and passenger mutations

Not every somatic mutation in a tumor drives cancer. A typical solid tumor contains hundreds to thousands of somatic mutations — the product of a lifetime of DNA replication errors, mutagen exposures, and repair failures. Most of these are passenger mutations: they happened to occur in the ancestor of the cancer cell but don't contribute to tumor growth.

A small minority are driver mutations: mutations that provide a growth, survival, or metastatic advantage to the cell carrying them. These are the mutations that natural selection within the tumor amplifies. Driver mutations are identified by their recurrence across tumors — when the same gene is mutated in 10%, 30%, or 70% of a given cancer type, it's almost certainly a driver.

The cancer gene census (COSMIC database, Wellcome Sanger Institute) catalogs over 700 genes with driver mutations across human cancers.

---

Oncogenes and tumor suppressors

Cancer drivers fall into two fundamental categories based on how their mutations work — which you already understand from Module 6.

Oncogenes are genes that, when mutated, are overactivated — producing a gain-of-function that drives uncontrolled cell growth. Normal proto-oncogenes regulate cell proliferation; a single dominant mutation can convert them into oncogenes.

Key oncogenes and their mechanisms:

• KRAS (mutated in ~25% of all cancers): RAS proteins are molecular switches that relay growth signals from cell surface receptors to the nucleus. KRAS mutations (most commonly G12D, G12V, G12C) lock the protein in a permanently active GTP-bound state, continuously signaling the cell to divide regardless of external signals. Long considered "undruggable" — the protein's smooth surface offers few binding sites. Sotorasib (2021, FDA-approved) was the first approved KRAS inhibitor, targeting specifically KRAS G12C.
• EGFR: Mutations in the epidermal growth factor receptor cause constitutive activation of downstream proliferation pathways. EGFR mutations are found in ~15% of non-small cell lung cancers (higher in Asian patients — ~50%). EGFR inhibitors (erlotinib, osimertinib) produce dramatic responses in EGFR-mutant tumors.
• MYC: A transcription factor that drives expression of hundreds of proliferation genes. MYC amplification (extra gene copies) is found in many cancers. Currently undruggable directly, but active research target.
• ERBB2 (HER2): Amplification causes overexpression of a growth factor receptor. HER2-amplified breast and gastric cancers respond to trastuzumab (Herceptin) — one of the earliest precision oncology successes.

Tumor suppressors are genes that normally constrain cell growth, repair DNA, or trigger apoptosis (programmed cell death). They require loss-of-function in both alleles to contribute to cancer (Knudson's two-hit hypothesis, Module 6). This means tumor suppressors typically act recessively within a cell.

Key tumor suppressors:

• TP53 (mutated in ~50% of all cancers): The "guardian of the genome." p53 is a transcription factor that activates DNA repair, cell cycle arrest, and apoptosis in response to cellular stress. When both copies are inactivated, cells with damaged DNA continue dividing instead of dying — accumulating further mutations. TP53 mutations are the most common alteration across all human cancers.
• BRCA1/2: DNA repair genes (homologous recombination). Germline mutations cause hereditary breast and ovarian cancer (Module 6). Somatic inactivation occurs in sporadic tumors. BRCA1/2-deficient tumors are sensitive to PARP inhibitors — drugs that exploit the DNA repair deficiency through synthetic lethality.
• RB1: The first tumor suppressor identified (retinoblastoma). pRb controls entry into cell division; its loss removes a key brake on proliferation.
• PTEN: A phosphatase that suppresses the PI3K/AKT growth signaling pathway. PTEN loss is common in prostate, endometrial, and glioblastoma.

---

Mutational signatures — the fingerprints of carcinogenesis

Every mutagen leaves a characteristic pattern of mutations. Because different types of DNA damage produce different mismatches, which are corrected (or not) in different ways, the type of carcinogen exposure a tumor experienced can often be inferred from the pattern of somatic mutations it contains.

These patterns are called mutational signatures, first systematically characterized by the Alexandrov et al. (2013) paper in Nature and maintained in the COSMIC database.

Examples of well-characterized signatures:

• Signature 4 (tobacco smoke): Predominantly C→A transversions, consistent with the DNA damage caused by polycyclic aromatic hydrocarbons in tobacco smoke. Found in lung cancers of smokers. A patient's lung cancer mutation spectrum can literally show whether they smoked.

• Signature 2 and 13 (APOBEC): C→T and C→G mutations in a specific sequence context, caused by APOBEC enzymes (cytidine deaminases that are part of the antiviral immune response but sometimes act on genomic DNA). Common in breast, bladder, and lung cancers.

• Signature 6 (mismatch repair deficiency): Predominantly insertions and deletions in repetitive sequences (microsatellites), caused by defective mismatch repair (MMR). Tumors with this signature — MSI-high (microsatellite instability-high) — are highly responsive to PD-1 checkpoint immunotherapy, regardless of tumor type. This led to the first tumor-agnostic FDA approval (pembrolizumab, 2017).

• Signature 3 (BRCA deficiency): Characterized by deletions at regions of microhomology, reflecting failed homologous recombination repair. Used to identify tumors with HR deficiency even when BRCA1/2 are not directly mutated.

Mutational signatures are now clinically useful: they can identify patients likely to respond to specific drugs, detect likely carcinogen exposures, and sometimes reveal hereditary cancer predisposition syndromes.

---

Tumor mutational burden and immunotherapy

Tumor mutational burden (TMB) is simply the number of somatic mutations per megabase of sequenced DNA in a tumor. High TMB tumors have many mutations; low TMB tumors have few.

Why does TMB matter? Because the immune system can potentially recognize neoantigens — novel peptides produced by somatic mutations that are presented on tumor cell surfaces via HLA molecules. The more mutations a tumor has, the more potential neoantigens it presents, and the more visible it may be to T cells.

Checkpoint immunotherapy — drugs that block PD-1, PD-L1, or CTLA-4 (immune checkpoints that normally suppress T cell activity) — releases the immune system to attack tumors. High-TMB tumors have more neoantigens for activated T cells to target, explaining why TMB predicts immunotherapy response across tumor types.

In 2020, the FDA approved pembrolizumab (Keytruda) for any solid tumor with TMB ≥ 10 mutations/megabase — the second tumor-agnostic approval, following the 2017 approval for MSI-high tumors. Tumor type doesn't matter; molecular profile does.

Melanoma and lung cancers exposed to carcinogens have the highest TMB; pediatric cancers and some liquid tumors have the lowest.

---

Clinical tumor sequencing — what actually happens

When a patient's tumor is sequenced clinically, the process and output are more complex than germline sequencing.

Matched tumor-normal sequencing: To distinguish somatic mutations from germline variants, clinical labs sequence both the tumor and a normal sample (usually blood) from the same patient. Variants present in the tumor but absent from blood are somatic; variants in both are germline. This matching is essential — otherwise, the germline variants that every person carries would appear as potential driver mutations.

Panel sequencing vs. WGS: Most clinical tumor sequencing uses a targeted gene panel — sequencing only a defined set of cancer-relevant genes (typically 300–500) at very high depth (1000x or more), rather than whole-genome sequencing. High depth is necessary to detect mutations present in only a fraction of tumor cells (subclonal mutations). Panels like FoundationOne CDx (FDA-cleared) and MSK-IMPACT are used clinically.

What a clinical tumor sequencing report contains:

• List of somatic mutations (SNVs, indels) in cancer-relevant genes
• Copy number alterations (amplifications, deletions)
• Structural variants (fusions)
• Microsatellite instability (MSI) status
• Tumor mutational burden (TMB)
• Actionable findings matched to FDA-approved therapies or clinical trials
• Germline incidental findings (depending on lab policy)

The report is interpreted by a molecular tumor board — a multidisciplinary team of oncologists, pathologists, geneticists, and bioinformaticians who collectively decide which findings are clinically actionable and how to use them.

---

Precision oncology — drugs matched to genomes

The defining promise of cancer genomics is precision oncology: matching drugs to the specific molecular alterations driving a patient's tumor rather than treating all cancers of the same anatomical origin the same way.

The most successful examples:

BCR-ABL / imatinib (Gleevec): The founding example. Chronic myelogenous leukemia (CML) is caused by the Philadelphia chromosome — a translocation between chromosomes 9 and 22 that creates the BCR-ABL fusion gene, encoding a constitutively active kinase. Imatinib (2001) is a small molecule inhibitor of the BCR-ABL kinase. CML went from a disease with a 5-year survival of ~30% to one where most patients achieve complete molecular remission on imatinib. This single drug transformed the natural history of a cancer.

EGFR mutations / EGFR inhibitors (lung cancer): EGFR-mutant NSCLC responds dramatically to EGFR tyrosine kinase inhibitors. Osimertinib (third-generation EGFR inhibitor) produces response rates of ~80% in EGFR-mutant tumors and has become standard first-line therapy. In EGFR wild-type tumors, EGFR inhibitors are essentially ineffective — demonstrating why genomic selection matters.

BRAF V600E / vemurafenib (melanoma): BRAF V600E is found in ~50% of melanomas. Vemurafenib, a BRAF inhibitor, produces response rates of ~50% in BRAF V600E-positive melanoma vs. near 0% in wild-type. Combining BRAF + MEK inhibition has further improved outcomes.

HER2 amplification / trastuzumab (breast and gastric cancer): Trastuzumab (Herceptin) targets the HER2 receptor, amplified in ~20% of breast cancers. In HER2-positive breast cancer, trastuzumab-based regimens reduce mortality substantially; in HER2-negative breast cancer, no benefit.

BRCA1/2 mutation / PARP inhibitors: BRCA1/2-deficient tumors rely on alternative DNA repair pathways, specifically PARP-mediated base excision repair. PARP inhibitors (olaparib, niraparib) block this backup pathway, causing lethal DNA damage accumulation specifically in HR-deficient cells — a concept called synthetic lethality. PARP inhibitors are approved for BRCA-mutant breast, ovarian, prostate, and pancreatic cancers.

---

Clonal evolution and drug resistance

Cancer is an evolutionary process. A tumor is not a genetically homogeneous mass — it is a population of cells with genetic diversity, competing for space and resources. This is intratumoral heterogeneity, and it is the primary driver of drug resistance.

When a targeted drug is applied, it creates a selection pressure: cells sensitive to the drug die; cells with mutations that confer resistance survive and expand. This is Darwinian evolution operating within a patient's body over weeks to months.

Mechanisms of acquired resistance:

• Secondary mutation in the drug target: The target evolves to evade the drug. In EGFR-mutant lung cancer, the T790M "gatekeeper" mutation confers resistance to first- and second-generation EGFR inhibitors — which is why osimertinib (designed to overcome T790M) replaced earlier drugs.
• Bypass pathway activation: The tumor activates a parallel signaling pathway that doesn't require the inhibited target. MET amplification bypasses EGFR inhibition in some lung cancers.
• Lineage plasticity: Tumor cells can change their identity. Some EGFR-mutant lung cancers transform into small-cell lung cancer under EGFR inhibitor pressure — a completely different histological type that is EGFR-independent.

Resistance is nearly inevitable for targeted monotherapy. The clinical strategies that have extended durability of response:

• Combination therapy: Targeting multiple nodes simultaneously makes it harder for resistance to evolve (analogous to HIV combination antiretroviral therapy)
• Sequential therapy: Use second- and third-generation drugs designed to overcome specific resistance mechanisms
• Monitoring for resistance emergence: Liquid biopsy enables early detection of resistance mutations before clinical progression

---

Liquid biopsy

A fundamental challenge in clinical oncology is monitoring tumor evolution without repeated invasive tissue biopsies. Liquid biopsy — analysis of tumor-derived materials in blood — has emerged as a transformative approach.

Circulating tumor DNA (ctDNA): Tumor cells shed fragments of DNA into the bloodstream. This cell-free DNA (cfDNA) contains somatic mutations characteristic of the tumor and can be detected in blood plasma using highly sensitive sequencing methods. ctDNA typically constitutes a small fraction of total cfDNA (0.01%–10%), requiring very high-sensitivity detection methods (digital PCR, error-corrected deep sequencing).

Clinical applications:

• Minimal residual disease (MRD) detection: After surgery or chemotherapy, ctDNA can detect residual tumor cells not visible on imaging — predicting relapse months before it would be clinically apparent. MRD-positive patients after curative-intent surgery have dramatically higher relapse rates; MRD negativity correlates with cure.
• Monitoring treatment response: ctDNA levels fall in responding tumors and rise in progressing ones — providing a molecular readout of treatment effect ahead of radiographic changes.
• Resistance mutation detection: When resistance mutations emerge, they appear in ctDNA before clinical progression, enabling early treatment switches.
• Early cancer detection: The most ambitious application — detecting cancer before symptoms appear from ctDNA in blood. Multi-cancer early detection (MCED) tests (Grail's Galleri test) can detect signal from 50+ cancer types from a blood draw. Current sensitivity for early-stage disease is limited (~40% for stage I, ~90% for stage IV), but the field is advancing rapidly.

---

Check yourself

1. A patient's NSCLC tumor is sequenced and found to have an EGFR exon 19 deletion (a common sensitizing mutation). They are started on osimertinib and respond well for 14 months. Their ctDNA then shows emergence of EGFR C797S, a mutation in the osimertinib binding site. What process explains the C797S emergence? What are the therapeutic options?

2. A tumor sequencing report shows: KRAS G12D mutation, TP53 loss (both alleles), high TMB (18 mutations/Mb), MSI-high status. Rank the actionable findings by immediacy of clinical implication and explain your reasoning.

3. A patient with ovarian cancer has a somatic BRCA2 frameshift in their tumor but no germline BRCA2 mutation. Their oncologist says PARP inhibitors won't work because "they only work if you inherited the BRCA2 mutation." Is this correct? Explain the biology.

4. A liquid biopsy test for early cancer detection has 85% sensitivity for stage III/IV disease and 30% sensitivity for stage I disease. A regulatory agency is deciding whether to approve it for population screening. What are the arguments for and against approval at this sensitivity level, and what additional data would you want?

---

Key facts to remember

• Driver vs. passenger mutations: drivers confer growth advantage and are recurrently mutated; passengers are incidental
• Oncogenes (GOF, dominant): KRAS (~25% of all cancers), EGFR, HER2, MYC
• Tumor suppressors (LOF, recessive within cell): TP53 (~50% of all cancers), BRCA1/2, RB1, PTEN
• Mutational signatures: carcinogen fingerprints in mutation patterns; Signature 4 = tobacco; MSI-high = MMR deficiency → immunotherapy responsive
• TMB ≥ 10 mut/Mb: FDA-approved pembrolizumab indication (2020); tumor-agnostic
• Matched tumor-normal sequencing distinguishes somatic from germline variants
• Synthetic lethality: BRCA deficiency + PARP inhibition = selective tumor cell death
• Clonal evolution drives acquired resistance; resistance is nearly inevitable with targeted monotherapy
• ctDNA: tumor DNA shed into blood; used for MRD detection, monitoring, resistance detection, early detection
• Galleri (Grail): MCED test, 50+ cancer types; ~30% sensitivity stage I, ~90% stage IV

---

Primary sources & references

• Alexandrov, L. B. et al. (2013). "Signatures of mutational processes in human cancer." Nature, 500, 415–421.
• Vogelstein, B. et al. (2013). "Cancer genome landscapes." Science, 339, 1546–1558.
• Druker, B. J. et al. (2001). "Efficacy and safety of a specific inhibitor of the BCR-ABL tyrosine kinase in chronic myeloid leukemia." NEJM, 344, 1031–1037.
• Mok, T. S. et al. (2009). "Gefitinib or carboplatin-paclitaxel in pulmonary adenocarcinoma." NEJM, 361, 947–957.
• Le, D. T. et al. (2017). "Mismatch repair deficiency predicts response to anti-PD-1 therapy across cancer types." Science, 357, 409–413.
• Ignatiadis, M. et al. (2021). "Liquid biopsy genotyping in advanced cancer." NEJM, 385, 2373–2374.
• Klein, E. A. et al. (2021). "Clinical validation of a targeted methylation-based multi-cancer early detection test." Annals of Oncology, 32, 1167–1177.