Precision Oncology

Precision oncology is the discipline of oncology that uses high-throughput molecular profiling—including hybrid-capture massive parallel sequencing, RNA-Seq fusion detection, immunohistochemistry, and circulating tumor DNA (ctDNA) liquid biopsy—to match the somatic genomic and immunologic landscape of an individual patient's tumor with targeted therapeutics. Whereas conventional cytotoxic chemotherapy targets all rapidly dividing cells by disrupting DNA synthesis or mitotic spindle assembly, precision oncology exploits tumor-specific somatic alterations through tyrosine kinase inhibitors, synthetic lethality and PARP inhibitors, antibody-drug conjugates, and immune checkpoint inhibitors.

The biological foundation of precision oncology rests on the principle of oncogene addiction, formulated by Bernard Weinstein in 2002. Although a malignant neoplasm accumulates hundreds to thousands of neutral passenger mutations due to genomic instability, tumor cell proliferation and survival remain dependent on one or a few dominant oncogenic driver alterations that rewire intracellular signaling networks. Acute pharmacological inhibition of the addicted driver pathway collapses survival signaling relative to pro-apoptotic stress signals, triggering rapid tumor regression while sparing normal host tissues that lack the somatic alteration.

Canonical Organ-Specific and Pathway-Matched Drivers

Across solid and hematologic malignancies, recurrent somatic point mutations, insertion–deletions (indels), copy-number amplifications, and chromosomal translocations define distinct molecular subsets with established targeted therapies:

• Chronic Myeloid Leukemia (BCR-ABL1): The Philadelphia chromosome translocation t(9;22)(q34;q11) generates the constitutively active BCR-ABL1 cytoplasmic tyrosine kinase, which is inhibited by ATP-competitive tyrosine kinase inhibitors (imatinib, dasatinib, nilotinib, ponatinib) or the allosteric myristoyl-pocket inhibitor asciminib.
• Non-Small Cell Lung Cancer (NSCLC): Adenocarcinomas of the lung are stratified into mutually exclusive driver subsets, including sensitizing EGFR exon 19 deletions and p.Leu858Arg (L858R) point mutations (treated with third-generation osimertinib), ALK rearrangements such as EML4-ALK (treated with alectinib, brigatinib, or lorlatinib), ROS1 fusions (entrectinib, repotrectinib), MET exon 14 skipping mutations (capmatinib, tepotinib), and EGFR exon 20 insertions (amivantamab).
• MAPK Pathway (BRAF V600E and KRAS G12C): In BRAF V600E-mutant melanoma, NSCLC, and anaplastic thyroid cancer, vertical dual blockade combining a BRAF inhibitor (dabrafenib or encorafenib) with a MEK1/2 inhibitor (trametinib or binimetinib) prevents paradoxical RAF dimer activation and delays resistance. For decades, RAS GTPases were considered undruggable due to their picomolar affinity for GTP/GDP and smooth surface topology; however, Kevan Shokat and colleagues identified a dynamic allosteric Switch-II pocket beneath the Switch-II loop of KRAS G12C, enabling covalent acrylamide inhibitors (sotorasib and adagrasib) to lock mutant Cys12 in its inactive GDP-bound state.
• Homologous Recombination Deficiency (BRCA1/2, PALB2): Germline or somatic loss-of-function mutations in homologous recombination DNA repair genes sensitize ovarian, breast, pancreatic, and castrate-resistant prostate cancers to PARP inhibitors (olaparib, niraparib, rucaparib, talazoparib).

Tumor-Agnostic (Tissue-Agnostic) Regulatory Approvals

Historically, oncology drug approvals were restricted to tumors arising from a specific anatomical organ (such as breast, colon, or lung). The discovery that certain genomic drivers and immunogenic biomarkers confer drug sensitivity regardless of tissue of origin led to tumor-agnostic (tissue-agnostic) FDA approvals, evaluated via histology-independent basket trials:

Evidence Tiers: OncoKB, AMP/ASCO/CAP, and ESCAT

Because comprehensive next-generation sequencing (NGS) panels (300 to 500+ genes) identify dozens of somatic variants per tumor, standardized evidence frameworks classify variants by clinical actionability to guide Molecular Tumor Boards (MTBs):

• OncoKB (Memorial Sloan Kettering): Assigns alterations to Level 1 (FDA-recognized biomarker predictive of response to an FDA-approved drug in that indication), Level 2 (standard care biomarker recommended by NCCN guidelines), Level 3A/3B (compelling clinical trial evidence in the same or another tumor type), Level 4 (compelling biological/preclinical evidence), and Levels R1/R2 (standard or investigational biomarkers predicting resistance to a therapy, such as KRAS/NRAS exons 2–4 mutations predicting resistance to anti-EGFR antibodies cetuximab and panitumumab in colorectal cancer).
• AMP/ASCO/CAP Consensus Tiers: Categorizes somatic variants into Tier I (variants of strong clinical significance, Level A and B), Tier II (variants of potential clinical significance, Level C and D), Tier III (variants of unknown clinical significance), and Tier IV (benign or likely benign variants).
• ESMO ESCAT: The European Society for Medical Oncology Scale for Clinical Actionability of molecular Targets ranks alterations from ESCAT Tier I (ready for routine clinical use) through Tier X (lack of evidence for actionability).

Circulating Tumor DNA (ctDNA) and Liquid Biopsy

Solid tumors shed fragmented cell-free DNA (cfDNA) into the bloodstream via apoptosis and necrosis, with nucleosome-protected fragments peaking around 166 base pairs (and ctDNA enriched in shorter 90–150 bp fragments). Ultra-deep error-corrected sequencing (using Unique Molecular Identifiers [UMIs] to suppress polymerase and sequencing artifacts down to variant allele fractions [VAF] of 0.01% to 0.1%) enables non-invasive liquid biopsy. Unlike a single-needle tissue biopsy that samples only one spatial region of a heterogeneous primary tumor, plasma ctDNA captures the aggregate mutational spectrum shed across primary and metastatic lesions, detects minimal residual disease (MRD) months before radiographic relapse on CT imaging, and monitors clonal evolution in real time during targeted therapy.

Under the Darwinian selective pressure of a targeted therapy, sensitive tumor clones are eliminated while pre-existing or newly acquired resistant subclones expand, typically leading to clinical progression within 9 to 36 months. Acquired resistance falls into three mechanistic classes:

1. On-Target Secondary Mutations: Somatic mutations in the drug-binding pocket sterically block inhibitor binding or increase ATP affinity while preserving catalytic activity. Examples include the EGFR p.Thr790Met (T790M) gatekeeper mutation against first-generation gefitinib/erlotinib, the EGFR p.Cys797Ser (C797S) mutation eliminating the covalent cysteine bond required by osimertinib, the ABL1 p.Thr315Ile (T315I) gatekeeper mutation against imatinib, and the ALK p.Gly1202Arg (G1202R) solvent-front mutation.
2. Off-Target Bypass Track Activation: The tumor reactivates downstream signaling (such as PI3K–AKT–mTOR or RAS–RAF–MEK–ERK) via an alternative receptor or pathway amplification independent of the inhibited target—most prominently focal MET amplification or ERBB2 amplification driving resistance to osimertinib in EGFR-mutant NSCLC, or secondary NRAS/MEK1 mutations in BRAF-treated melanoma.
3. Histologic Lineage Plasticity (Phenotypic Transdifferentiation): Rather than mutating the target pathway, the tumor Epigenetically reprograms its cellular identity into a lineage that no longer expresses or depends on the original driver. In 5% to 15% of EGFR-mutant lung adenocarcinomas treated with osimertinib (and in androgen-receptor-inhibited prostate adenocarcinomas), concomitant loss of the tumor suppressors RB1 and TP53 permits transdifferentiation into small-cell neuroendocrine carcinoma, which downregulates EGFR protein expression and requires platinum–etoposide chemotherapy.

Why do tumors with KRAS mutations fail to respond to anti-EGFR antibodies in colorectal cancer?

EGFR is a transmembrane receptor tyrosine kinase positioned upstream of the intracellular RAS–RAF–MEK–ERK signaling cascade. When KRAS or NRAS harbors an activating mutation in codon 12, 13, 61, 117, or 146, the RAS protein signals constitutively downstream regardless of whether upstream EGFR is blocked by cetuximab or panitumumab, making wild-type RAS confirmation mandatory prior to anti-EGFR therapy.

What is the clinical difference between somatic and germline testing in precision oncology?

Somatic testing sequences DNA or RNA extracted from tumor tissue or plasma ctDNA to identify acquired mutations present only in the neoplastic clone (such as EGFR L858R or KRAS G12C). Germline testing sequences normal host leukocytes or saliva to detect inherited cancer predisposition syndromes (such as BRCA1/2, Lynch syndrome mismatch repair genes, or TP53 Li-Fraumeni syndrome) that affect both targeted therapy eligibility and family cascade screening.

How does a liquid biopsy distinguish true tumor mutations from clonal hematopoiesis (CHIP)?

In aging individuals, benign expansion of hematopoietic stem cell clones carrying somatic mutations in DNMT3A, TET2, ASXL1, or occasionally TP53, ATM, and CHEK2—termed Clonal Hematopoiesis of Indeterminate Potential (CHIP)—releases mutant DNA from white blood cells into plasma cfDNA. Paired sequencing of the buffy coat (matching peripheral blood leukocytes) alongside plasma cfDNA subtracts hematopoietic CHIP variants from true tumor-derived ctDNA.

1. Weinstein, I. B. (2002). "Cancer: Addiction to oncogenes—the Achilles heal of cancer." Science, 297(5578), 63–64. https://doi.org/10.1126/science.1073096
2. Chakravarty, D., Gao, J., Phillips, S. M., et al. (2017). "OncoKB: A precision oncology knowledge base." JCO Precision Oncology, 1, 1–16. https://doi.org/10.1200/PO.17.00011
3. Ostrem, J. M., Peters, U., Sos, M. L., Wells, J. A., & Shokat, K. M. (2013). "K-Ras(G12C) inhibitors allosterically control GTP affinity and effector interactions." Nature, 503(7477), 548–551. https://doi.org/10.1038/nature12796
4. Wan, J. C. M., Massie, C., Garcia-Corbacho, J., et al. (2017). "Liquid biopsies come of age: towards implementation of circulating tumour DNA." Nature Reviews Cancer, 17(4), 223–238. https://doi.org/10.1038/nrc.2017.7