Tumor Markers: What They Are and What They Cannot Tell You

Tumor markers are biological molecules — measured in blood, urine, or tissue — that rise in the presence of certain cancers or in response to cancer. They are used to support cancer diagnosis, guide treatment, monitor whether therapy is working, and detect recurrence after treatment ends. Understanding tumor markers is essential for any cancer patient — because the most common mistake with these tests is over-interpreting a single result.

No tumor marker is cancer-specific. Every marker used in clinical practice has non-cancer causes of elevation. A single elevated tumor marker is a clinical signal that warrants follow-up — not a cancer diagnosis. Tissue biopsy is required to confirm cancer, regardless of what tumor markers show. For the full landscape of cancer blood tests beyond protein markers, see our guide to the cancer blood test.

What Tumor Markers Are — and Are Not

Tumor markers are proteins, antigens, hormones, enzymes, or genetic material produced either directly by cancer cells or by the body in response to a tumor. They are measurable in blood, urine, tissue samples, or other body fluids.

Over 200 biological molecules have been studied as potential tumor markers. Only a fraction are validated and used in routine clinical practice. Tumor markers serve three primary roles:

1. Diagnostic support. When cancer is clinically suspected — based on symptoms, imaging, or physical exam — tumor markers help narrow the differential. Example: AFP and beta-hCG with a testicular mass.
2. Treatment monitoring. Serial measurements assess whether cancer is responding to therapy. Declining marker = tumor cell death. Rising marker = inadequate response or progression.
3. Surveillance for recurrence. After curative treatment, periodic tumor marker measurement detects early recurrence — often weeks to months before imaging or symptoms.

Tumor Markers — Complete Reference Table

Tumor Markers Deep Dive — The Most Ordered Tests

PSA — What the Number Really Means

PSA is prostate-organ specific, not cancer-specific. It is produced by all prostate tissue — normal, enlarged, inflamed, or cancerous. Many patients assume that PSA above 4.0 ng/mL means prostate cancer, and below 4.0 ng/mL means it’s absent. Neither is true.

In the “gray zone” of PSA 4–10 ng/mL, approximately 25–30% of men will have prostate cancer on biopsy — meaning 70–75% will not. Roughly 25% of prostate cancers occur in men with PSA below 4.0 ng/mL. To refine risk beyond a simple PSA level, clinicians use:

• Free/total PSA ratio: Lower free fraction (<10–15%) is more suspicious for cancer vs. BPH
• PSA density: PSA ÷ prostate volume (MRI-based); >0.15 ng/mL/cc raises concern
• 4K score: Combines total PSA, free PSA, intact PSA, and hK2 kallikrein into an aggressive cancer probability
• phi (Prostate Health Index): FDA-cleared; outperforms PSA alone in the 4–10 gray zone

Post-treatment: After radical prostatectomy, PSA should fall to undetectable (<0.1 ng/mL). Two consecutive PSA readings >0.2 ng/mL define biochemical recurrence. After radiation, the Phoenix criteria define biochemical failure as PSA nadir + 2 ng/mL.

Screening controversy: USPSTF (2018) gives PSA screening Grade C for men aged 55–69 — individualized shared decision-making, not a population-wide recommendation. The American Cancer Society recommends annual PSA discussion starting at 50 (average risk), 45 (African American men or first-degree relative <65), and 40 (multiple first-degree relatives).

CA-125 — Why It Is Not an Ovarian Cancer Screening Test

CA-125 is elevated in approximately 80% of epithelial ovarian cancers at diagnosis — but in only about 50% of stage I disease, when treatment is most curative. This early-stage insensitivity is one reason CA-125 fails as a general population screening test.

The other reason is poor specificity: endometriosis, fibroids, PID, pregnancy, liver disease, and menstruation all elevate CA-125, especially in premenopausal women. Both the PLCO Cancer Screening Trial (U.S.) and UKCTOCS trial (UK) confirmed that CA-125-based screening did not reduce ovarian cancer mortality in average-risk women. USPSTF gives a Grade D recommendation for CA-125 screening.

CEA — The Colorectal Cancer Follow-Up Marker

CEA is not sensitive enough for colorectal cancer screening — elevated in only about 25% of stage I CRC. Its primary role is post-operative surveillance for stages II–III CRC: ASCO and NCCN guidelines recommend CEA every 3–6 months for 2–3 years after surgery, then every 6 months for another 2 years.

The clinical value is a 2–3 month lead time over imaging for detecting hepatic or pulmonary metastases — enough time to enable earlier surgical intervention for oligometastatic disease. Heavy smokers have baseline CEA elevation into the 5–10 ng/mL range; normal limits differ (<2.5 ng/mL non-smokers, <5.0 ng/mL smokers).

AFP — Liver Cancer and Testicular Cancer

AFP is produced by the fetal liver and yolk sac; levels are very low in healthy adults. In hepatocellular carcinoma (HCC) surveillance, AFP is measured every 6 months combined with liver ultrasound in high-risk patients (cirrhosis, chronic HBV). AFP >400 ng/mL with cirrhosis was historically considered diagnostic of HCC; current AASLD/EASL guidelines require imaging confirmation regardless of AFP level.

In testicular cancer, AFP is a critical marker for non-seminomatous germ cell tumors (NSGCT). Critical rule: pure seminoma does NOT produce AFP. If AFP is elevated with apparent pure seminoma on pathology, non-seminomatous elements must be assumed and NSGCT protocols followed. AFP half-life is 5–7 days; post-orchiectomy AFP normalization is tracked to confirm adequate treatment response.

CA 19-9 — The Pancreatic Marker That Cannot Screen

CA 19-9 has two important biological limitations. First, approximately 5–10% of the population is Lewis antigen-negative — genetically unable to produce CA 19-9, meaning this marker will always read zero regardless of cancer burden. Second, biliary obstruction from any cause (gallstones, cholangitis, PSC) raises CA 19-9 to thousands of U/mL — making it impossible to distinguish pancreatic cancer from benign biliary disease on CA 19-9 alone.

USPSTF Grade D for pancreatic cancer screening applies to CA 19-9 as well as all other screening approaches. CA 19-9 is most useful for monitoring response to chemotherapy in known pancreatic cancer and evaluating resectability.

Thyroglobulin — Only After Thyroid Surgery

Thyroglobulin as a tumor marker is meaningless unless all thyroid tissue has been removed. After total thyroidectomy with or without RAI ablation, any detectable thyroglobulin indicates residual or recurrent differentiated thyroid cancer.

Anti-thyroglobulin antibodies (TgAb) are present in 20–25% of thyroid cancer patients and can falsely lower thyroglobulin immunoassay results. Every thyroglobulin measurement must include TgAb testing. In TgAb-positive patients, the TgAb trend serves as a surrogate marker for tumor burden. Stimulated thyroglobulin <0.1 ng/mL in TgAb-negative patients = excellent response.

Chromogranin A — The PPI Confounder

CgA is a secretory protein from neuroendocrine cells, useful for monitoring carcinoid tumors, pheochromocytoma, and gastroenteropancreatic neuroendocrine tumors (GEP-NETs). The most common cause of an unexpected elevated CgA in clinical practice is PPI (proton pump inhibitor) use — PPIs cause gastric ECL cell hyperplasia that can raise CgA 2–10× above baseline. PPI should be held for 2 weeks before testing when possible.

CgA assays are not standardized across laboratories. Serial monitoring must use the same assay at the same reference lab to be meaningful.

Molecular and Genomic Tumor Markers

Beyond protein blood markers, molecular tumor markers measured in tumor tissue or via liquid biopsy are central to modern cancer management. They fall into two categories:

Predictive markers (guide which therapy to use):

• EGFR mutations, ALK/ROS1 fusions (NSCLC) → tyrosine kinase inhibitor selection
• KRAS/NRAS/BRAF V600E (colorectal cancer) → anti-EGFR antibody eligibility
• HER2 amplification (breast, gastric) → trastuzumab/pertuzumab eligibility
• BRCA1/2 mutations (breast, ovarian, prostate, pancreatic) → PARP inhibitor eligibility
• MSI-H/dMMR (any tumor) → pembrolizumab (Keytruda) eligibility, tumor-agnostic
• NTRK fusions (any tumor) → larotrectinib/entrectinib eligibility

Prognostic markers (predict outcome):

• Ki-67 proliferation index (breast cancer): >30% = high-grade aggressive biology
• Oncotype DX Recurrence Score (ER+ breast cancer): RS ≥26 = chemotherapy benefit in premenopausal women (TAILORx trial)
• Decipher Prostate: genomic classifier for post-prostatectomy metastatic risk

Liquid biopsy tests (ctDNA from a blood draw) can detect some of these mutations when tissue is unavailable or when monitoring for acquired resistance during treatment. For a complete overview, see the cancer blood test article.

When Specific Tumor Markers Are Ordered

How to Interpret a Tumor Marker Result

A single elevated tumor marker is the beginning of an investigation, not the end of one. Before acting on any result:

1. Rule out pre-analytical errors. Hemolysis falsely elevates NSE. PPI use falsely elevates CgA. Anti-Tg antibodies falsely lower thyroglobulin. A lab assay change can produce apparent “rises” that reflect methodology, not tumor biology.
2. Assess the trend, not the single value. A stable elevated marker unchanged over 12 months has very different significance from one that has doubled in 3 months.
3. Correlate with imaging. Most protocols require imaging (CT, PET-CT, MRI) before clinical action on a rising marker. Treatment decisions depend on where the relapse is.
4. Confirm the marker is appropriate for your cancer type. Off-label marker use has no validated interpretation and generates clinical anxiety without actionable information.

If your tumor marker result has come back and your next step is a follow-up appointment, our guide to what to expect at a cancer checkup can help you prepare. And if you want to understand your baseline cancer risk before a screening discussion, our cancer risk assessment guide explains the key risk factors and tools clinicians use.

Tumor Markers and Cancer Screening — What’s Approved

PSA is the only protein tumor marker with population-level screening protocols — and even PSA has a Grade C USPSTF recommendation for men 55–69 (individual shared decision-making). Every other major protein tumor marker lacks USPSTF A or B screening endorsement. CA-125, CA 19-9, CEA, AFP, and CgA screening are not recommended for average-risk adults because false-positive cascades outweigh benefits in low-prevalence populations.

For validated, guideline-backed cancer screening by age and risk, see our guide to preventive cancer screening.

Frequently Asked Questions

This article is for educational purposes only and does not constitute medical advice. Tumor marker results should be interpreted by a qualified healthcare provider in the context of your full clinical picture, imaging, and history.