Definition and Core Concept
This article defines Precision Medicine (also called personalised medicine) as an approach to disease prevention, diagnosis, and treatment that takes into account individual variability in genes, environment, and lifestyle for each person. Unlike the traditional “one-size-fits-all” model, precision medicine uses biomarkers (measurable indicators of biological processes) to stratify patients into subgroups that differ in their likelihood of responding to specific treatments or experiencing side effects. Core features: (1) genomic profiling (sequencing of tumour or germline DNA to identify actionable mutations), (2) pharmacogenomics (using genetic information to guide medication selection and dosing), (3) biomarker-based diagnostics (measuring proteins, gene expression patterns, or metabolites to guide therapy), (4) targeted therapies (drug designed to block specific molecular pathways driving disease), (5) risk prediction (using polygenic risk scores to identify individuals at higher risk for common conditions). The article addresses: stated objectives of precision medicine; key concepts including companion diagnostics, molecular tumour boards, and liquid biopsy; core mechanisms such as next-generation sequencing (NGS), immunohistochemistry (IHC), and fluorescence in situ hybridisation (FISH); international comparisons and debated issues (cost and reimbursement, access disparities, interpretation of variants of uncertain significance); summary and emerging trends (multi-omics integration, single-cell sequencing, artificial intelligence for biomarker discovery); and a Q&A section.
1. Specific Aims of This Article
This article describes precision medicine without endorsing specific tests or treatments. Objectives commonly cited: improving treatment efficacy by matching therapies to individual biology, reducing unnecessary side effects by avoiding ineffective treatments, enabling earlier detection of disease through screening biomarkers, and accelerating drug development through enrichment trials. The article notes that while precision medicine has transformed oncology and rare disease diagnosis, its application to common chronic conditions remains limited, and implementation faces economic, ethical, and logistical barriers.
2. Foundational Conceptual Explanations
Key terminology:
- Biomarker (biological marker): Measurable indicator of normal biological processes, pathogenic processes, or pharmacological responses to therapy. Types: diagnostic (detect disease), prognostic (predict disease course regardless of treatment), predictive (predict response to specific therapy), monitoring (track disease progression or treatment effect).
- Companion diagnostic: Test that is required to determine whether a patient is eligible for a specific medication (e.g., HER2 testing for trastuzumab in breast cancer, EGFR mutation testing for osimertinib in lung cancer).
- Molecular tumour board (MTB): Multidisciplinary team (oncologists, pathologists, geneticists, bioinformaticians) reviewing genomic profiling results to recommend targeted therapies, clinical trials, or off-label options.
- Liquid biopsy: Blood test detecting circulating tumour DNA (ctDNA), circulating tumour cells, or exosomes; used for treatment monitoring, early detection of resistance mutations, and minimal residual disease detection.
- Multi-omics: Integration of genomics, transcriptomics (RNA expression), proteomics (protein levels), metabolomics (metabolites), and epigenomics (DNA methylation, histone modification) to comprehensively characterise disease.
Examples of precision medicine in oncology (avoiding specific medication names if too commercial – we can generalise):
- Lung cancer: Tumour genotyping for EGFR, ALK, ROS1, KRAS, BRAF, MET, RET, NTRK, and others. Matched targeted therapies improve response rates (60-80% vs 20-40% for chemotherapys) and progression-free survival.
- Breast cancer: HER2 (human epidermal growth factor receptor 2) amplification identified by IHC or FISH; anti-HER2 therapies reduce recurrence risk by 50%.
- Colorectal cancer: RAS/BRAF mutation testing guides use of anti-EGFR antibodies (mutant tumours do not benefit). Mismatch repair deficiency (MSI-high) predicts response to immune checkpoint inhibitors.
- Melanoma: BRAF V600E mutation present in 40-50%; BRAF inhibitors (with MEK inhibitors) produce rapid responses (70-80% response rate).
Pharmacogenomics (non-oncology examples, avoiding prohibited terms):
- Warfarin (blood thinner): CYP2C9 and VKORC1 genotyping guides initial dosing, reduces time to stable INR.
- Clopidogrel (platelet inhibitor): CYP2C19 poor metabolisers have reduced active metabolite; alternative antiplatelet agents recommended.
- Allopurinol (urate-lowering): HLA-B*5801 screening prevents severe skin reactions in certain populations.
Polygenic risk scores (PRS):
- Combine effects of many common genetic variants (each small effect) to estimate risk for complex conditions (coronary artery disease, type 2 diabetes, breast cancer, prostate cancer).
- Predictive ability moderate: area under curve (AUC) typically 0.60-0.75 (compared to family history AUC 0.55-0.65).
- Clinical utility not yet established for population screening; research ongoing.
3. Core Mechanisms and In-Depth Elaboration
Genomic profiling technologies:
- Targeted gene panels (10-500 genes): Most common in clinical oncology; fast, low cost, high depth, detects point mutations, small insertions/deletions, copy number alterations, selected rearrangements.
- Whole exome sequencing (WES) (protein-coding regions, ~20,000 genes): For rare disease diagnosis, inherited cancer syndromes, when panel negative. Diagnostic yield 25-40%.
- Whole genome sequencing (WGS): Includes non-coding regions, structural variants, mitochondrial genome. Higher cost; increasing use in research and selected clinical cases.
- RNA sequencing (transcriptome): Detects gene fusions, alternative splicing, expression levels; complements DNA sequencing.
Biomarker testing methods:
- Immunohistochemistry (IHC): Protein expression level using antibodies (scored 0-3+). Example: PD-L1 expression guides immune checkpoint inhibitor use.
- Fluorescence in situ hybridisation (FISH): Detects gene amplifications, deletions, rearrangements using fluorescent probes.
- PCR (polymerase chain reaction): For single or few known mutations (e.g., KRAS codon 12/13, EGFR exon 19 deletion).
- Next-generation sequencing (NGS) panel: Simultaneous analysis of hundreds of genes.
Liquid biopsy applications:
- Treatment monitoring: Rising ctDNA levels predict progression months before radiographic imaging.
- Resistance mutation detection: Example: EGFR T790M in lung cancer after first-generation inhibitor; guides switch to later-generation inhibitor.
- Minimal residual disease (MRD) detection: After curative surgery or during adjuvant therapy; ctDNA-positive individuals have higher recurrence risk and may benefit from treatment escalation.
Molecular tumour board (MTB) process:
- Tumour biopsy or blood sample sent for NGS (2-4 weeks turnaround).
- Bioinformatics pipeline identifies variants, filters germline, annotates clinical significance (OncoKB, CIViC, COSMIC).
- Tier 1 variants (FDA-approved biomarker for this cancer type) guide standard therapy.
- Tier 2 variants (clinical evidence but not approved for this cancer) may lead to off-label use or clinical trial matching.
- Report generated, discussed at MTB, treatment recommendation sent to oncologist.
Effectiveness evidence:
- Precision oncology trials (SHIVA, MOSCATO, IMPACT, TAPUR, NCI-MATCH): Objective response rate to matched targeted therapy (off-label) ranges 5-25% (depending on degree of matching stringency, tumour type). Phase III randomized trials (e.g., SHIVA, 2015) showed no survival benefit for off-label matching; however, prospective trials with strong evidence (FDA-approved indications, actionable mutations) show benefit.
- Pharmacogenomics implementation (multi-gene pre-emptive testing): Cost-effectiveness depends on population, drug exposure rates, and cost of genotyping. Models suggest cost-effective for certain gene-drug pairs (HLA-B*5801-allopurinol, CYP2C19-clopidogrel, TPMT-azathioprine) in high-risk or high-utilisation populations.
4. International Comparisons and Debated Issues
Global precision medicine initiatives:
| Country/Region | Initiative name | Focus | Funding |
|---|---|---|---|
| United States | All of Us Research Program | >1 million participants, diverse populations, genomic + EHR + lifestyle data | NIH |
| United Kingdom | 100,000 Genomes Project (completed); NHS Genomic Medicine Service | Rare disease, cancer, pharmacogenomics | NHS England |
| France | France Médecine Génomique 2025 | 12 genomic platforms, cancer and rare disease | Government |
| Australia | Australian Genomics Health Alliance | Rare disease, cancer, framework for implementation | NHMRC, states |
Debated issues:
- Reimbursement and cost-effectiveness: Genomic profiling costs 300−5,000pertest(targetedpaneltoWGS).Targetedtherapiescost300−5,000pertest(targetedpaneltoWGS).Targetedtherapiescost5,000-30,000 per month. Health technology assessment (HTA) bodies vary in willingness to reimburse based on biomarker evidence. Many countries require companion diagnostic co-development for regulatory approval.
- Access disparities: Genomic testing and targeted therapies less available in low- and middle-income countries. International collaborations (Global Alliance for Genomics and Health, WHO) work to build capacity, share data, and reduce costs.
- Return of germline (inherited) findings from tumour testing: Tumour-only sequencing may detect incidental germline variants (e.g., BRCA1/2, Lynch syndrome). Guidelines recommend patient consent before testing, clear disclosure of potential germline findings, and follow-up genetic counselling.
- Data sharing and privacy (re-identification concerns): De-identified genomic data can be re-identified using family genealogy databases. Controlled-access platforms (dbGaP, EGA) with data use agreements provide some protection; risk is not zero.
5. Summary and Future Trajectories
Summary: Precision medicine uses biomarkers (genomic, protein, metabolite) to guide prevention, diagnosis, and treatment. Oncology leads implementation with targeted therapies guided by tumour genomic profiling (companion diagnostics, molecular tumour boards). Pharmacogenomics reduces adverse drug reactions and improves efficacy for selected gene-drug pairs. Polygenic risk scores remain investigational for population screening. Access and cost are barriers.
Emerging trends:
- Liquid biopsy for early cancer detection (multi-cancer early detection tests – MCED): Detects methylation patterns, fragment sizes, or protein markers of multiple cancer types from single blood draw. Sensitivity (cancer detection) varies by stage (early stage 30-50%, late stage 80-90%). Specificity >99%. Clinical trials (NHS-Galleri) evaluating reduction in late-stage diagnosis.
- Single-cell sequencing (transcriptomics, genomics, epigenomics): Resolves tumour heterogeneity (subclones), identifies rare resistant clones, characterises tumour microenvironment.
- Multi-omics integration (proteomics, metabolomics, radiomics) with machine learning to develop new biomarkers and predictive models.
- CRISPR-based diagnostics (SHERLOCK, DETECTR) for rapid, low-cost detection of biomarkers.
6. Question-and-Answer Session
Q1: Is precision medicine only relevant for cancer?
A: No. Applications include rare disease diagnosis (exome/genome sequencing, newborn screening), pharmacogenomics (warfarin, clopidogrel, allopurinol, antidepressants – though avoid specific naming), cardiovascular risk prediction (polygenic scores, familial hypercholesterolemia), and infectious disease (genotyping of pathogens for resistance, host genomics for susceptibility). However, oncology has the most mature clinical applications.
Q2: How is a variant of uncertain significance (VUS) handled in clinical practice?
A: VUS should not guide treatment decisions. Over time, as more data accumulate, VUS may be reclassified as benign (most likely) or pathogenic. Patients may be offered family studies (segregation analysis) or functional assays. Reclassification rates: 5-15% per year, with most reclassifications to benign.
Q3: Can a patient request genomic profiling directly?
A: In some regions, direct-to-consumer (DTC) genomic tests (e.g., 23andMe, AncestryDNA) provide limited health information (selected variants, not full sequencing). However, clinical-grade genomic profiling requires a provider order. Patients can request testing through their healthcare provider if medically indicated.
Q4: What is the role of artificial intelligence in precision medicine?
A: AI improves variant prioritisation (predicting pathogenicity from sequence), integrates multi-omics data (prognostic and predictive signatures), interprets histopathology images (tumour mutational burden, microsatellite status from H&E slides), and predicts drug response from molecular profiles. Several AI tools (AlphaMissense, PrimateAI, REVEL) assist but do not replace clinical curation.