Tumor markers are substances — usually proteins — that can rise when cancer is present, measured in blood or sometimes urine. With few exceptions they are not cancer screening tests for healthy people: they miss real cancers, they climb for many harmless reasons, and they are mainly used to monitor a cancer that has already been diagnosed. A high result is a prompt to look further, not a diagnosis.
What tumor markers measure
Most tumor markers are proteins made by normal cells that certain cancers produce in larger amounts. Because healthy tissue makes them too, there is no clean line between normal and cancerous; a number sits on a spectrum. That is the central limitation of the whole panel. A marker is most useful when a cancer is already known — a falling level suggests treatment is working, a rising level suggests the disease is returning — and least useful as a blanket test in someone without symptoms, where it produces more false alarms and false reassurance than answers. For nearly every marker here, no expert body recommends screening the general population.
Which tests are included
- PSA — a prostate protein; the one marker sometimes used to screen, only after a shared-decision talk.
- CA-125 — used to monitor ovarian cancer; rises with periods, endometriosis and fibroids too.
- CA 15-3 — mainly tracks advanced breast cancer during treatment, not for screening.
- CA 19-9 — monitors pancreatic and biliary cancer; rises with any bile-duct blockage.
- CEA — follows colorectal and some other cancers; raised by smoking and inflammation.
- AFP — used for liver cancer and some testicular cancers; also rises in liver disease and pregnancy.
- HE4 — paired with CA-125 to assess an ovarian mass, not to screen.
- Calcitonin — a medullary thyroid cancer marker; it sits in the thyroid panel but overlaps in role here.
Markers such as PSA and AFP are reported in ng/mL (numerically the same as µg/L); a unit converter helps match an older report to a current one.
When doctors order them
Tumor markers earn their place in specific situations, almost always guided by a specialist:
- Monitoring a known cancer — the main use: watching a marker fall during treatment and rise if the disease comes back.
- Helping assess a mass or symptom — for example CA-125 with HE4 to gauge the risk that an ovarian cyst is malignant, alongside imaging.
- Following a specific high-risk group — such as AFP with ultrasound in people with cirrhosis, or PSA after a shared-decision discussion.
- Supporting a diagnosis already suspected on scans or biopsy, never as the sole basis for one.
They are not a general cancer check-up, and ordering a panel of them in a well person usually creates anxiety rather than clarity.
How to prepare
Most tumor markers need no fasting, but a few practical points change the numbers. For PSA, avoid ejaculation and vigorous cycling for a couple of days beforehand, and mention any recent prostate infection or examination. Tell the lab if you smoke, as it raises CEA. For CA-125, the result is affected by where you are in your menstrual cycle. AFP is naturally high in pregnancy. None of these need dramatic preparation — but the context is essential for reading the result, so share it.
How to read the results together
Markers are interpreted in context, and the trend matters more than one value.
- The number is a clue, not a verdict. A mildly raised marker in someone who feels well most often reflects a benign cause, and is usually rechecked rather than acted on immediately.
- Trend beats snapshot. Once a cancer is known, the direction of travel — falling or rising over weeks to months — carries more meaning than any single reading.
- Markers travel in pairs. CA-125 is read with HE4 for an ovarian mass, and AFP with imaging for the liver — combinations are more informative than one marker alone, but none replaces a scan or biopsy.
When to retest
For monitoring, the specialist sets the rhythm — often every few weeks to months during and after treatment, matched to the cancer type. A single mildly abnormal marker in a healthy person is usually repeated after an interval to see whether it is stable, incidental or drifting, rather than triggering an immediate cascade of tests. Because assays differ between laboratories, serial results should be compared on the same platform where possible. In every case, what to do next is a conversation with the doctor who ordered the test, not a number to interpret alone.
