Precision Medicine in Cancer: Biomarkers, Targeted Therapy, and Genetic Testing

A clear guide to precision oncology — how tumours are profiled, what biomarkers mean, how targeted therapy differs from chemotherapy and immunotherapy, what companion diagnostics are, and what the limits of the approach are.

Precision Medicine in Cancer: Biomarkers, Targeted Therapy, and Genetic Testing

Modern cancer treatment has moved well beyond the era of treating all cancers of a given organ the same way. For a growing number of cancer types, the molecular profile of the tumour — its specific genetic mutations, protein expression patterns, and other biological features — determines which treatments are most likely to work.

This field is called precision oncology or precision medicine. It has produced remarkable advances in some cancer types and is actively changing others. This guide explains how it works, what the key terms mean, and what its current limits are.


What Precision Medicine Means in Practice

Traditional oncology classified cancers by the organ where they arose (lung cancer, breast cancer, colon cancer) and by how the cells looked under a microscope (histology). Treatment decisions followed accordingly.

Precision oncology adds a molecular layer: it asks what specific changes in this tumour are driving its growth? The answer might be a mutation in a specific gene, an overabundance of a particular protein, a gene fusion, or a specific pattern of instability in the tumour’s DNA.

When a molecular target is identified and a drug exists that reliably inhibits it, treatment can be matched to the tumour’s biology rather than just its organ of origin. This can mean substantially better outcomes for patients whose tumours carry an actionable target — and avoids exposing patients to drugs unlikely to help them.

Not every tumour has an actionable target. Not every target yet has an effective drug. Precision medicine is transforming some cancers more than others.


How Tumours Are Profiled

Tissue biopsy and NGS

Next-generation sequencing (NGS) of tumour tissue is the primary tool for molecular profiling in clinical oncology. A biopsy is needed to obtain tumour material. The laboratory analyses the DNA (and sometimes RNA) of the tumour cells, looking for:

  • Somatic mutations — changes in the tumour DNA not present in normal cells (acquired, not inherited)
  • Copy number alterations — sections of the genome that are amplified or deleted
  • Gene fusions — where parts of two different genes join together to create an abnormal driving protein (e.g., ALK or ROS1 fusions in lung cancer)
  • Microsatellite instability (MSI) — a pattern of errors in the tumour’s DNA repair machinery
  • Tumour mutational burden (TMB) — the total number of mutations in the tumour, which can predict response to immunotherapy in some contexts

Liquid biopsy

Some profiling tests analyse circulating tumour DNA (ctDNA) from a blood sample. These liquid biopsies are less invasive and can be useful when tissue is insufficient or hard to obtain. However, they have limitations: sensitivity is lower than tissue NGS for detecting some alterations, and negative results do not rule out the presence of a mutation.

Germline versus somatic testing

An important distinction:

  • Somatic testing analyses mutations that arose in the tumour itself (not inherited, not present in every cell of the body). This guides treatment decisions for the current cancer.
  • Germline testing analyses inherited mutations present in all cells. Results have implications for the patient’s relatives and may qualify the patient for certain treatments (e.g., PARP inhibitors for BRCA1/2 germline mutations). It also has implications for genetic counselling.

For some cancers — including pancreatic, ovarian, and breast cancer — both types of testing are recommended.


Key Biomarkers in Oncology

KRAS

Cancers: Pancreatic (>90%), colorectal (~40–50%), lung adenocarcinoma (~25–30%)

KRAS mutations lock a cell-signalling switch permanently “on,” driving uncontrolled cell growth. KRAS was historically undruggable. The first KRAS-targeting drugs — sotorasib and adagrasib — received FDA approval for KRAS G12C-mutated lung cancer in 2021–2022. In May 2026, Phase 3 results published in the New England Journal of Medicine showed that daraxonrasib, a multi-selective RAS(ON) inhibitor, roughly doubled survival compared with chemotherapy in previously treated metastatic pancreatic cancer — a significant shift in a difficult disease.

See: KRAS Mutations Explained: Why Some Cancer Targets Are Hard to Treat

EGFR (Epidermal Growth Factor Receptor)

Cancers: Lung adenocarcinoma (~10–15% of Western patients; ~40–50% in East Asian patients), some others

EGFR mutations drive tumour growth through a receptor tyrosine kinase signalling pathway. Several oral drugs (gefitinib, erlotinib, afatinib, osimertinib) inhibit mutated EGFR, and osimertinib has become standard first-line treatment for EGFR-mutated advanced lung cancer, substantially improving outcomes over chemotherapy.

HER2 (ERBB2)

Cancers: Breast (~15–20%), gastric/gastro-oesophageal (~15–20%), lung (HER2 mutations, ~3%), colorectal (HER2 amplification, ~3–5%)

HER2 overexpression or amplification drives tumour growth through another receptor tyrosine kinase. Targeted agents (trastuzumab, pertuzumab, lapatinib, trastuzumab deruxtecan) have transformed outcomes in HER2-positive breast and gastric cancer. Trastuzumab deruxtecan (T-DXd) — an antibody-drug conjugate — has also shown activity in HER2-mutated lung and other solid tumours, representing a “pan-tumour” HER2 approach.

BRCA1 and BRCA2

Cancers: Breast, ovarian (most commonly), also pancreatic, prostate, and others when germline

BRCA proteins are involved in repairing double-strand breaks in DNA. Tumours with BRCA1/2 mutations are unable to repair DNA efficiently via homologous recombination — a vulnerability that PARP inhibitors (olaparib, niraparib, rucaparib) exploit. PARP inhibitors have been approved for BRCA-mutated ovarian, breast, pancreatic, and prostate cancers. Germline BRCA testing has implications for family members’ cancer risk.

MSI-H / Mismatch Repair Deficiency (dMMR)

Cancers: Colorectal (~15% of early-stage, ~4–5% metastatic), endometrial (~25–30%), and rarer fractions of other tumour types

When the mismatch repair (MMR) system fails, tumour DNA accumulates thousands of small insertion/deletion errors — a state called microsatellite instability-high (MSI-H). These tumours have a very high mutational burden, generating many abnormal proteins recognised by the immune system. This makes them highly responsive to immune checkpoint inhibitors (pembrolizumab). Pembrolizumab is FDA-approved for any MSI-H solid tumour regardless of origin — one of the first truly tumour-agnostic approvals. Testing for MMR protein expression (IHC) or MSI by PCR or NGS is straightforward and recommended for all newly diagnosed colorectal cancers and increasingly for other tumours.

ALK and ROS1 Fusions

Cancers: Lung adenocarcinoma (ALK ~3–5%, ROS1 ~1–2%)

Gene fusions involving ALK or ROS1 produce abnormal kinase proteins that drive tumour growth. Multiple approved ALK inhibitors exist (crizotinib, ceritinib, alectinib, brigatinib, lorlatinib), with later-generation drugs offering better central nervous system penetration and overcoming resistance to earlier drugs. These cancers tend to occur in younger, non-smoking patients and respond well to targeted therapy.

NTRK Fusions

Cancers: Rare; found in <1% of most solid tumours but in high proportions of certain rare tumour types (infantile fibrosarcoma, secretory breast carcinoma)

NTRK gene fusions produce TRK proteins that drive growth. Two oral TRK inhibitors — larotrectinib and entrectinib — are FDA-approved for any solid tumour with an NTRK fusion regardless of site. Response rates in NTRK fusion-positive tumours have been high. This is a paradigmatic example of a tumour-agnostic targeted approval.

PD-L1

PD-L1 expression on tumour cells is a biomarker for some immune checkpoint inhibitor treatments, but it behaves differently from the mutation-based biomarkers above. PD-L1 is not a driver mutation — it reflects the tumour’s interaction with the immune system. Testing is predictive (though imperfectly) for response to PD-1/PD-L1 checkpoint inhibitors (pembrolizumab, nivolumab, atezolizumab). Note that PD-L1 positive does not guarantee response, and PD-L1 negative does not guarantee non-response — MSI-H status is more predictive than PD-L1 in many settings.


Targeted Therapy vs Immunotherapy

These two categories are distinct, though both are sometimes labelled “modern” or “non-chemo” treatments.

Targeted TherapyImmunotherapy (Checkpoint Inhibitors)
How it worksBlocks a specific cancer-driving molecule (a mutated protein, receptor, or kinase)Removes brakes on the patient’s immune T cells by blocking PD-1, PD-L1, or CTLA-4
Requires specific mutationUsually yes — the specific target must be presentNot necessarily — though MSI-H/TMB are predictive biomarkers
Example drugsOsimertinib (EGFR), trastuzumab (HER2), olaparib (BRCA)Pembrolizumab, nivolumab, ipilimumab
Key side effectsOn-target toxicities (rash, diarrhoea, cardiac effects — depends on target)Immune-related adverse events (colitis, pneumonitis, endocrinopathies, hepatitis)
Cancer “addiction” requiredYes — the tumour must depend on the target being inhibitedNo — activating the immune response can work across targets

Companion Diagnostics

A companion diagnostic (CDx) is a test that has been approved alongside a specific drug, where use of the drug requires (or strongly recommends) confirming the relevant biomarker before prescribing.

Examples:

  • EGFR mutation testing is required before prescribing osimertinib for lung cancer
  • HER2 testing (IHC/FISH) is required before prescribing trastuzumab for breast or gastric cancer
  • BRCA1/2 testing is required before prescribing olaparib for certain cancer types

The FDA now co-approves many targeted drugs with their companion diagnostic. This ensures that expensive and potentially toxic drugs are given only to patients likely to benefit.


Limitations of Precision Medicine

Precision oncology has real limits that are important to understand.

Not all tumours have actionable targets. Across all solid tumours, molecular profiling finds an actionable alteration in roughly 30–40% of patients — and whether an approved drug exists for that alteration varies. For pancreatic cancer, outside of the small subgroups with BRCA mutations or MSI-H status, the dominant KRAS driver has only recently shown signs of being druggable, and no approved targeted therapy currently exists for most patients.

Resistance almost always develops. Targeted therapy rarely eliminates cancer permanently. Cancer cells evolve mechanisms to bypass the blocked pathway — secondary mutations, pathway reactivation, or phenotypic change. Managing resistance often means switching to a different drug targeting the same pathway (e.g., later-generation EGFR inhibitors) or combining drugs.

Tissue availability and quality can be limiting. Comprehensive NGS requires sufficient tumour tissue of adequate quality. Some tumours are difficult to biopsy, or biopsy specimens may be too small or poorly preserved for molecular testing.

Cost and access vary widely. Comprehensive NGS and approved targeted therapies are expensive. In many health systems, access depends on insurance, national reimbursement decisions, and geography. Not all patients globally have equal access to molecular profiling or targeted drugs.

Tumour heterogeneity. A single biopsy may not capture all the molecular changes present across a tumour or its metastases. Different parts of the tumour, or metastatic deposits, can have different molecular profiles. A targeted drug may work on cells with the target but fail against cells that have already evolved alternative drivers.


What Happens After Molecular Profiling?

Once NGS results are available, an oncologist reviews them to determine:

  • Are there alterations for which approved drugs exist?
  • Are there alterations that suggest eligibility for clinical trials?
  • Does the profile suggest resistance to certain treatments?

In centres with molecular tumour boards, a multidisciplinary team reviews complex cases. For many patients, profiling confirms that standard chemotherapy is still the appropriate backbone, with targeted treatment added or held in reserve. For a smaller group, profiling reveals an actionable finding that changes the treatment plan.


FAQ

Q: What is precision medicine? An approach to cancer treatment that uses the molecular characteristics of the tumour to guide treatment choices — matching therapy to specific genetic or molecular changes rather than treating all cancers of a given type identically.

Q: What is a biomarker? A measurable biological feature — a gene mutation, protein level, or molecular pattern — that provides information about the cancer’s biology or likely response to treatment.

Q: How is my tumour profiled? Usually through next-generation sequencing (NGS) of tissue from a biopsy. This analyses hundreds of cancer-related genes simultaneously to look for mutations, fusions, copy number changes, and other alterations.

Q: What is the difference between targeted therapy and chemotherapy? Chemotherapy kills rapidly dividing cells broadly. Targeted therapies are designed to block a specific molecular change driving the cancer. Targeted therapies generally only work when the specific target is present in the tumour.

Q: What is the difference between targeted therapy and immunotherapy? Targeted therapy blocks a specific cancer-driving molecule. Immunotherapy (checkpoint inhibitors) works by releasing immune suppression, allowing the patient’s own T cells to attack the cancer. Both are distinct from standard chemotherapy.

Q: Does everyone benefit from precision medicine? Not yet. Precision oncology is most transformative in cancers where well-defined driver mutations exist and effective drugs have been developed (e.g., EGFR-mutated lung cancer, HER2-positive breast cancer). In many cancers, actionable targets remain limited.

Q: What is a companion diagnostic? A regulatory-approved test required before prescribing a specific targeted drug — to confirm that the relevant biomarker is present. It ensures that the drug is used only in populations likely to benefit.


Further Reading



Educational only — not a substitute for professional medical advice.