Tumour Markers: Types, Uses, Limitations and Clinical Interpretation

Tumour markers are measurable substances or molecular features associated with cancer. They may be produced by tumour cells, released because of tumour-related tissue injury, or expressed by normal cells in response to a tumour. Markers can support diagnosis, estimate prognosis, select treatment and monitor response, but most are neither sensitive nor specific enough to diagnose cancer alone or to screen the general population.

Learning outcomes

The learner should be able to define tumour markers, classify circulating and tissue biomarkers, explain their clinical uses and limitations, recognise common markers and associated conditions, interpret trends safely and avoid inappropriate screening or diagnosis based on an isolated result.

1. Definition and classification

1.1 Circulating or fluid markers

These are measured in blood, urine, cerebrospinal fluid, pleural fluid or another specimen. Examples include alpha-fetoprotein (AFP), beta-human chorionic gonadotropin (β-hCG), prostate-specific antigen (PSA), carcinoembryonic antigen (CEA) and CA-125.

1.2 Tissue markers

These are detected in biopsy or surgical tissue. Immunohistochemistry identifies proteins such as oestrogen receptor, progesterone receptor, HER2, mismatch-repair proteins and PD-L1. Flow cytometry characterises haematological malignancies, while cytogenetics and sequencing detect rearrangements or mutations.

1.3 Molecular biomarkers

DNA, RNA, methylation, gene-fusion, copy-number and circulating tumour-DNA features can classify a tumour or identify a target. A biomarker can be diagnostic, prognostic, predictive or a combination.

2. What tumour markers can do

Clinical use	How the marker helps	What it cannot do alone
Support diagnosis	Raises or lowers suspicion in a compatible clinical picture	Usually cannot establish cancer without tissue or definitive disease-specific evidence
Risk or surveillance	Monitors selected high-risk patients, such as AFP in chronic liver disease pathways	Does not replace imaging or specialist follow-up
Prognosis	Associates with tumour burden, subtype or expected behaviour	Cannot determine an individual outcome with certainty
Predictive treatment selection	Identifies a target or pathway likely to respond to a therapy	A marker does not guarantee response; resistance can develop
Response monitoring	Serial fall may support treatment response; rise may signal progression or relapse	Early changes can reflect treatment-related cell death, inflammation or assay variation
Recurrence surveillance	Trend after definitive treatment may prompt imaging or review	Not every recurrence raises a marker and not every rise is recurrence

3. Why isolated results mislead

• False positive: benign inflammation, pregnancy, liver or kidney disease, smoking, menstruation, prostatitis, assay interference or another cancer raises the marker.
• False negative: the tumour does not produce the marker, the disease burden is small, the specimen is poorly timed or the assay is insensitive.
• Biological variation: levels fluctuate with physiology, hydration, organ clearance and tumour kinetics.
• Analytical variation: different laboratories, platforms and reference ranges are not always interchangeable.
• Lead-time and overdiagnosis: finding a marker earlier does not automatically improve survival or justify treatment.

A marker result should be interpreted as a probability modifier. Ask: “What is the pre-test probability, what benign causes are plausible, and what confirmatory test will change management?”

4. Common circulating tumour markers

Marker	Common associations	Benign/non-cancer causes and cautions	Typical clinical use
AFP	Hepatocellular carcinoma; non-seminomatous germ-cell tumours	Pregnancy, chronic hepatitis, cirrhosis and liver injury; some cancers do not raise AFP	Surveillance/diagnostic support in selected liver-risk pathways; germ-cell tumour assessment
β-hCG	Gestational trophoblastic disease; some germ-cell tumours	Pregnancy, pituitary hCG, assay interference and rare tumours	Diagnosis, staging and response monitoring with imaging and pathology
PSA	Prostate cancer	Benign prostatic enlargement, prostatitis, urinary retention, ejaculation and instrumentation	Risk assessment, selected screening after shared decision-making, treatment/recurrence monitoring
CEA	Colorectal and some pancreatic, gastric, breast and lung cancers	Smoking, inflammation, liver disease and benign gastrointestinal disease	Baseline and serial monitoring, especially after colorectal cancer treatment
CA-125	Epithelial ovarian and related cancers	Menstruation, pregnancy, endometriosis, pelvic inflammation, liver disease and effusions	Monitoring selected ovarian cancers; not a stand-alone population screening test
CA 19-9	Pancreatic and biliary cancers	Cholestasis, cholangitis, pancreatitis and Lewis-antigen non-expression	Monitoring in a compatible diagnosis; not diagnostic alone
CA 15-3/CA 27-29	Breast cancer and some other adenocarcinomas	Benign breast/liver disease and other cancers	Selected treatment/recurrence monitoring; not general screening
Calcitonin	Medullary thyroid carcinoma	Renal disease, other neuroendocrine conditions and assay variation	Diagnosis and surveillance in selected patients/families
Thyroglobulin	Differentiated thyroid cancer after thyroid tissue removal/ablation	Residual benign thyroid tissue and anti-thyroglobulin antibodies	Recurrence surveillance with antibody interpretation
LDH	Lymphoma, leukaemia, germ-cell tumours and high tumour burden	Tissue injury, haemolysis, liver disease, infection and muscle injury	Prognostic/risk stratification support, not tumour-specific diagnosis

5. Tissue and molecular biomarkers

5.1 Breast and other solid tumours

• Oestrogen and progesterone receptors: predict potential benefit from endocrine therapy.
• HER2: overexpression or amplification may support HER2-targeted treatment; test methods and thresholds matter.
• PD-L1: may inform immunotherapy in selected cancers, but scoring is tumour-specific.
• Mismatch-repair proteins/microsatellite instability: identify repair deficiency and may predict immunotherapy response or inherited Lynch syndrome risk.

5.2 Haematological malignancy

Flow cytometry, immunophenotyping, cytogenetics and molecular tests distinguish leukaemia and lymphoma lineages, classify risk and detect measurable residual disease. Results must be interpreted with blood film, marrow morphology and clinical findings.

5.3 Actionable genetic biomarkers

Specific mutations or fusions may guide targeted therapy in lung, colorectal, melanoma, thyroid, breast and other cancers. A negative result may reflect tumour heterogeneity, inadequate tissue or assay limitations. A tissue biomarker is usually a somatic finding unless germline testing confirms inherited risk.

6. Interpreting a tumour-marker result

1. Confirm the indication: Why was it ordered—symptoms, known cancer, high-risk surveillance or treatment planning?
2. Review pre-test probability: Age, symptoms, examination, imaging, family history, exposures and prior cancer.
3. Check the specimen and method: timing, fasting or pregnancy status where relevant, laboratory reference range and assay platform.
4. Look for benign explanations: infection, inflammation, obstruction, organ failure, smoking, menstruation, pregnancy or recent instrumentation.
5. Use serial trends: compare the same assay where possible; consider half-life and expected treatment kinetics.
6. Confirm with imaging, pathology or specialist review: do not treat an isolated number as a diagnosis.

7. Kinetics and serial monitoring

After tumour treatment, a marker may fall according to its biological half-life. A slow decline, plateau or second rise can suggest residual or recurrent disease, but timing matters. Surgical manipulation, chemotherapy-related cell lysis, infection and organ dysfunction can cause transient changes. The marker should be interpreted alongside symptoms, examination and imaging.

Use the same laboratory method whenever possible. Record dates, treatments, pregnancy status, organ function and relevant procedures. A small change near the assay’s analytical variation may not be clinically meaningful.

8. Screening limitations

Most circulating tumour markers do not work well for screening asymptomatic people because they miss cancers and produce false positives. Screening tests must demonstrate benefit, acceptable harms and a clear pathway for confirmatory diagnosis. Established screening programmes—such as cervical, breast or colorectal pathways—should not be replaced by unvalidated panels.

Selected markers have a role in high-risk surveillance or a specific clinical context. For example, AFP may be used within an appropriate chronic-liver-disease surveillance programme, and PSA decisions require shared decision-making that considers age, risk, benefits and harms.

9. Marker-specific pitfalls

AFP

AFP can rise in pregnancy and active hepatitis or cirrhosis. A normal AFP does not exclude hepatocellular carcinoma, and a high result does not prove it. Combine with liver imaging and specialist assessment.

β-hCG

Always consider pregnancy before interpreting β-hCG. Germ-cell tumours may produce β-hCG or AFP; seminoma classically does not produce AFP. False-positive low-level results can occur from assay interference or pituitary production.

PSA

PSA is prostate-specific, not cancer-specific. Benign enlargement, prostatitis, retention, ejaculation and instrumentation can alter values. Interpret age, prostate volume, symptoms, examination and serial change according to local guidance.

CA-125 and CA 19-9

Pelvic inflammation, menstruation, endometriosis, cholestasis, cholangitis and pancreatitis can raise these markers. A raised result requires clinical correlation and imaging, not automatic cancer labelling.

CEA

Smoking and benign liver or bowel inflammation can elevate CEA. It is most useful as a baseline and trend in a known, marker-producing cancer rather than as a general screening test.

10. Tumour markers in emergencies

11. Case applications

Case 1: Raised CA 19-9 with jaundice

A patient with obstructive jaundice has a high CA 19-9. Cholestasis and cholangitis can elevate the marker, so manage the obstruction/infection and obtain appropriate imaging. Reassess after resolution rather than diagnosing pancreatic cancer from one value.

Case 2: AFP in chronic hepatitis

A patient with chronic hepatitis B has a rising AFP. The result increases concern but does not prove hepatocellular carcinoma. Follow the liver-surveillance pathway with ultrasound or cross-sectional imaging and specialist review.

Case 3: PSA after instrumentation

A PSA is measured shortly after urinary retention and catheterisation. These events can affect PSA. Record timing, treat the acute problem and repeat or investigate according to local prostate guidance rather than making an immediate cancer diagnosis.

12. Reporting and communication

• State the marker, value, units, reference range, assay method and date.
• Explain what the result can and cannot establish.
• Use “abnormal marker” rather than “cancer” until the diagnosis is confirmed.
• Give a clear plan for confirmatory testing, referral and repeat measurement.
• Discuss anxiety, false positives and the possibility of a normal marker in cancer.

References and further reading

• NCI: Tumour markers fact sheet.
• NCI: Tumour marker tests in common use.
• MedlinePlus: Tumour marker tests.
• NCI: Biomarker testing for cancer treatment.
• Follow local oncology, laboratory, screening and referral guidelines; biopsy and multidisciplinary assessment remain central to cancer diagnosis.