Cancer Biomarkers

Protein Tumour Markers

• PSA, CA-1225, CEa and Ca19-9

• Classic serum markers measured in labs

• Useful for monitoring known disease rather than first-line diagnosis

Oncofetal markers

• AFP, β-hCG

• Important in germ-cell tumours, while AFP also has role in hepatocellular carcinoma pathways

Tumour burden/turnover

• LDH, urate, calcium, serum free light chains

• Don’t usually diagnose tumour directly, but can reflect tumour mass, aggressive disease or treatment-related complications

Genomic biomarkers

• Tissue or in blood

• Actionable mutations, resistance mutations, molecular evidence of minimal residual disease

• Central for precision oncology

Sensitivity, specificity, and why prevalence matters

Low-prevalence settings such as screening asymptomatic populations, even a reasonably sensitive and specific test may produce far more false positives than true positives. This is one reason why very few serum tumour markers are suitable for population screening

• Sensitivity = how well a test detects disease when disease is present

• Specificity = how well a test remains negative when disease is absent

• Positive predictive value depends on prevalence, low prevalence setting = positive result is more likely to be false

• Neg pred value also influenced by prevalance, but normal markers do not exclude cancer

• Tumours are not stand alone rule in or rule out tools - need to include symptoms, imaging, clinical context, pre-test probability

When are tumour markers actually useful?

• Histology and staging best for diagnosis

• Biomarkers are for prognosis, treatment monitoring and recurrence detection

• Screening = limited role bc poor specificity > large numbers of false positives, outcome benefit for applications not shown

• Diagnosis = support suspicions, e.g. liver lesion w elevated AFP, but tissue diangosis and imaging criteria is most definitive

• Prognosis and risk stratification = baseline levels can correlate w tumour burden or outcome

• Monitoring = serial measurements are most useful, trends over time may reveal response to treatment or recurrence

• SIngle values WEAK, trends STRONG

Pre-analytical factors that can mislead

• Timing = markers can rise after treatment bc of tumour lysis, inflammation, tissue injury; means post-treatment rise not always indicative of progression

• Sample quality = haemolysis, lipaemia, delayed processing cna delay assays

• Physiological and benign conditions = pregnancy, menstruation, inflammation and renal impairment can alter marker concentrations

• Clinical procedures = Prostate manipulation can raise PSA, cholestasis can raise CA19-9

• DO NOT OVER INTERPRET!!!!!!!

Analytical Pitfalls

• if result is implausible, do not take at face value

• Heterophile antibody interference = false positive immunoassays. Relevant for hCG; if result doesnt fit clinical picture, may need follow up labs e.g. urine testing, alternative platforms, interference-blocking approaches

• High-dose hook effect = extremely high analyte concentrations saturate assay system, give paradoxically false low result - dilution and reassay may reveal true concentration

• Assay drift + changes in method can complicate monitoring, serial-follow up is best done on same platofrm

• EXAM POINT!!! lab should be contacted in result is implausible

Biological Variation and Trend Interpretation

• Small tumour marker changes may reflect biological/analytical variation, not true disease progression.

• A rise is only meaningful if it exceeds expected variation.

• Always interpret trends with clinical context:

  • surgery

  • infection/inflammation

  • chemotherapy

  • assay changes

  • sampling interval

• Trend + clinical context is better than single isolated value.

Interpreting Tumour Markers

1. Define the clinical question, is the test being sued for screening, diagnositc support, prognosis, or monitoring?

2. Assess pre-test probability, consider benign differentials

3. Assay context and possible interference?

4. Next clinical step?

Questions to keep in mind

• Why was this marker ordered?

• Does the patient have a known malignancy, symptoms, or imaging findings?

• Could infection, inflammation, pregnancy, renal disease, or obstruction explain the result?

• Does the result need repeating, confirming on another platform, or interpreting as a serial trend?

• What does this result change next: repeat, image, biopsy, refer, or alter treatment?

PSA

• Not cancer-specific

• Prostate-associated, can be raised in neign prostatic hyperplasia, prostatis, UTI, prostate manipulation - cancer is only a possible explanation

• Risk marker rather than diagnositc - requires further labs e.g. MRI, biopsy, adge-adjusted thresholds, PSA density and velocity, free to total PSA

• Repeat timing and appropriate referral must be considered

CA-125 and HE4

• CA-125 associated w ovarian cancer, but lacks specifcity - rises in endometriosis, fibroids, menstruation, pregnancy, inflammation. Must have symptoms and imaging (pelvic ultrasound)

• HE4 and tools such as ROMA can improve specificity but local practice can vary

• Useful in symptomatic pathways and monitoring rather than population screening

AFP

• Hepatocellular carcinoma - must be interpreted carefully bc chronic hepatitis + cirrhosis can cause rises, should he paired w ultrasounds and etc.

• Can also be related to germ cell tumours

Germ cell tumour markers β-hCG, AFP and LDH

• β-hCG + AFP for diagnosis, staging and monitoring of testicular germ cell cancer

• LDH less specific but can reflect tumour burden and prognosis

• HOWEVER pos hCG is not automatic for cancer bc of heterophile antibodies

• Need urine hCG, repeat testing on alternative platofrm, heterophile blocking reagents

• DO NOT LEAP STRAIGHT FROM POS HCG = GERM CELL TUMOUR, consider ASSAY INTERFERENCE

CEA

• Colorectal cancer, mostly for monitoring, useful for recurrence monitoring

• Not specific - smoking, liver disease, inflammation + other malignancies, must see an upward trend

CA19-9

• Pancreatic, billiary malignancy, but can be confused by cholestasis and benign biliary obstruction + pancreatitis

• May fall after biliary decompression even if cancer still present

• Some Lewis-antigen individuals do not produce CA19-9 at all

• Need imaging and tissue diagnosis

Thyroid markers

Thyrobloublin for monitoring thyroid cancer after thyroidectomy + radioiodine treatment

• Anti-thyroglobulin antibodies can invalidate results

• Calcitonin is important in medullary thryoid cancer, can correlate w tumour burden

Haematological Cancer markers

• More diagnosis as opposed to solid tumours

• Serum free light chains + immunofixation important in plasma cell disorders e.g. myeloma

• β2-microglobulin and albumin > prognosis, LDH > aggressiveness

• Tumour lysis synrdome risk assessmen: high urate, phosphate, portassium, creatnine and LDH w low calcium > dangerous metaboloic complications

Tissue Matters

• Tumour tupe, grade, margins and microenvironmnetal infrmation

• Biomarkers are not replacements for pathology, they compliment each other in a multidisciplinary team

Liquid Biopsy

• Analysis of tumour-derived material in blood, particularly tumour DNA, cells and other components circulating

• Minimally invasive and repeatable

• Protein tumour markers can lack specificty and lag behind tumour evolution, liquid biopsy can show mutations, resistance and minimal resudal disease

• Does not replace histology

ctDNA

Circulating tumour DNA, fraction of cell-free DNA in blood from tumour cells

• amoutn depends on tumour burden, biology and how much tumour shedding

• Negative ctDNA does not equal no cancer

• Clonal haematopoiesis of indeterminate potential (CHIP) can produce variants in blood-derived DNA, mimic tumour associated mutations

Liquid Biopsy Technologies

• PCR + digital droplet PCR - highly sensitive for known variants, cannot discover new abnormalities

• Targeted NGS panels = actionable and resistance mutations, but sensitivity depends on depth and deisgn

• Whole genome/exome sequencing = wide coverage, but low sensitivity and higher complexitiy

• Methylations and fragmentomics = tissue of origin, but need validation

Clinical Use of Liquid Biopsy

• Genotyping when tissue is limited/unsafe to obtain - ctDNA can find actionable mutations, support treatment selection

• Minimal resudal disease detecting - post treatment dtDNA can mean residual disease and higher recurrence risk

• Minotirng tumour evolution + resistance during therapy - emerging mutations and changing biology

• Negative liquid biopsy DOES NOT PROVE ABSENCE OF MUTATIONS OR DISEASE

Early detection liquid biopsies

• Early cancers shed very little ctDNA so sensitivity fails

• Screeening needs high specificity otherwise ppl will go thorugh harmful follow up tests

• Tissue of origin prediction is based on probability and not perfect, can hinder investigation

• Ethical and systematic issues of overdiagnosis innit

Liquid biopsy vs traditional diagnosis

• Minimally invasive, repeatable and captures hetergeneity better than single tissue sample and can detect recurrence

• But has false negatives, low tumour fractions, artefacts and CHIP

• Trad is still good bc of histology, grading, morphology and microenvironment - good for diangosis and planning management

• Good to combine all thse things according to clinical question

Practical issues of biomarjers

• Good systems are needed, Tube choice, transport time, processing, assay selection, report wording and MDT protocols needed

• For liquid biopsy, delayed processing can increase background cell-free DNA and reduce sensitivity

• All reports need to state detection limits, genes covered, tumour fractions limits and interpretive cautions

• Governance and ethics are important - incidental findings e.g. germline variants or clonal haematopoiesis need pathways

• Quality assurance, validation, audit and equity are needed

Interprofessional collaboration

• Biomarkers not to be used by one professional group

• Assay governance, test stewardship, interpretive comments and communicating is importnat for cirtical results

• Oncology and surgery define clinial question and act on result

• Pathology and molecular labs intergrate tissue and genomic findings

• Radiology and endoscopy localise disease and provide tissue safely

• Biomarker testing should be embedded in agreed MDT pathways rather than used ad hoc