Comprehensive Guide to Biomarkers in Personalised Healthcare

A biomarker is a characteristic that is objectively measured and evaluated as an indicator of three specific states:
- Normal biological processes.
- Pathogenic processes.
- Pharmacologic responses to a therapeutic intervention.
Biomarkers are analyzed using instruments such as a biosensor.

Biomarkers can manifest in various biological or clinical forms, including:
- Genes.
- Proteins.
- Imaging results.
- Clinical measures.
Biomarkers can be utilized as standalone indicators or in combination as part of a complex biomarker signature.
Specific examples of biomarker modalities include:
- Electrophysiological: such as $EEG$ and $ECG$ .
- Physiological: including Blood Pressure ( $BP$ ), Heart Rate ( $HR$ ), and Forced Expiratory Volume ( $FEV$ ).
- Biological and Physical: Behavioural, histological, and cytological measures.
- Imaging: Medical imaging techniques such as $MRI$ and $CT$ .
- Omics analysis: Comprehensive biological datasets including genomics, transcriptomics, proteomics, and metabolomics.

Biomarkers serve four distinct roles in the context of personalised healthcare, though a single biomarker may serve more than one function:
1. Diagnostic biomarkers: Utilized for the early detection and classification of diseases.
2. Predictive biomarkers: Utilized to identify patients likely to develop a disease, predict responses to a specific pharmacotherapy, or identify the likelihood of experiencing an adverse effect.
3. Metabolism biomarkers: Utilized for defining or monitoring the appropriate dosage of a drug.
4. Outcome biomarkers: Utilized to forecast the progression of a disease or the likelihood of recurrence.

Diagnostic biomarkers enable the detection of a condition at an early stage, which can include presymptomatic and predictive testing.
Presymptomatic testing involves assessing whether a patient has inherited an autosomal dominant gene variant that causes a late-onset condition before any clinical signs or symptoms manifest.
The primary purposes of this testing are:
- To prevent the disease or its progression by identifying patients who require early treatment or lifestyle modification while they remain asymptomatic.
- To provide critical information that allows for future planning.
Examples of conditions appropriate for presymptomatic testing include:
- Huntington disease.
- Familial hypercholesterolaemia.
- Familial adenomatous polyposis.
- Lynch syndrome.
- Hereditary breast and ovarian cancer.
- Cervical cancer.

Biomarkers are used as predictive or prenatal testing tools within families to identify individuals who are heterozygous for a recessive allele.
When homozygous, these recessive alleles cause a genetic disorder or an altered phenotype.
This process is commonly known as carrier testing.
Examples of monogenic inherited diseases for which carrier testing is used include:
- Cystic fibrosis.
- Duchenne muscular dystrophy.
- Fragile X syndrome.
- Haemophilia A and B.

Biomarkers are employed for disease classification when a patient becomes symptomatic to confirm or rule out specific causes of illness.
These tools are essential for guiding clinical treatment decisions.
Specific examples include:
- Inflammation markers: Used to detect inflammatory responses.
- Cardiac troponin I assessment: Used for the diagnosis of myocardial infarction.
- Breast cancer type determination: Utilizing $BRCA1$ and $BRCA2$ mutations, $HER-2$ overexpression, and oestrogen receptor positive characterisation.
- Detection of infectious agents: Identifying pathogens such as $HIV$ , $H.\,Pylori$ , and $TB$ .

These biomarkers define or monitor the response to pharmacotherapy to achieve maximum efficacy and minimize Adverse Drug Reactions ( $ADRs$ ).
Examples of metabolism-based monitoring include:
- Warfarin: Assessment through $CYP2C9$ genotyping and International Normalized Ratio ( $INR$ ) measurement.
- Irinotecan: Evaluation of uridine diphosphate glucuronosyltransferase activity.
- Thiopurines: Evaluation of thiopurine $S$ -methyltransferase activity.

Predictive biomarkers determine which patients will benefit from a particular therapy in terms of efficacy or the avoidance of $ADRs$ .
In some instances, testing must occur before a pharmacotherapy can be prescribed; these are known as companion diagnostics.
These biomarkers may overlap with other categories:
- Outcome biomarkers: Because they can forecast therapeutic response.
- Metabolism biomarkers: When the biomarker in question is a drug-metabolising enzyme.
Clinical examples of personalized drug treatments include:
- Tamoxifen: Dependent on the presence of oestrogen receptors in breast cancer biopsy.
- Trastuzumab: Dependent on $HER-2$ overexpression in breast cancer biopsy.
- Gefitinib: Dependent on $EGFR$ mutation in non-small cell lung cancer.
- Gleevac: Dependent on the Philadelphia chromosome in positive chronic myeloid leukaemia.
- Abacavir: Screening for the $HLA$ - $B5701$ allele, which is associated with severe hypersensitivity reactions.
- Simvastatin: Screening for solute carrier organic transporter variation associated with statin-induced myopathy.

Outcome tools are used to predict or monitor the progression, prognosis, and recurrence of a disease, which directly informs treatment choices.
These applications often rely on biomarker signatures.
Examples of prognostic tools include:
- Breast cancer prognosis: EndoPredict, Oncotype DX breast recurrence score, and Prosigna.
- Inflammatory Bowel Disease ( $IBD$ ): PredictSure-IBD.

Biomarker signatures are unique, specific, and complex combinations of two or more biomarkers that reflect a particular biological state.
Signatures offer higher specificity and precision compared to individual biomarkers or general patterns.
The combination of multiple markers results in greater predictive power than any single marker used in isolation.
High-level illustrative example of a signature involving two biomarkers:
- Biomarker 1: Genetic variation (represented as Wildtype or $Allele\,I$ ).
- Biomarker 2: Level of expression of an $mRNA$ (represented as High or Low).
Outcome Matrix for the Signature:
- If Biomarker 1 is Wildtype and Biomarker 2 is High, the prognosis is Good.
- If Biomarker 1 is Wildtype and Biomarker 2 is Low, the prognosis is Good.
- If Biomarker 1 is $Allele\,I$ and Biomarker 2 is High, the prognosis is Poor, requiring aggressive treatment and monitoring.
- If Biomarker 1 is $Allele\,I$ and Biomarker 2 is Low, the prognosis is Good.