Tumor Marker Tests: What They Actually Tell You (And What They Don't)

A tumor marker test is a blood test that measures a substance — usually a protein — that some cancers release into the bloodstream. It is one of the most requested items in any health screening panel, and one of the most misread. So here is the short version before anything else: a raised tumor marker is not a cancer diagnosis, and a normal tumor marker is not an all-clear. Both directions fail often enough that no serious medical body recommends these tests to screen healthy people for cancer.

We say that as a company that sells them. They sit inside our screening panels, and we think that placement is defensible — but only with the sentence above attached. This guide explains what each common marker is, what raises it for entirely harmless reasons, what the major guideline bodies actually say, and where these tests genuinely earn their keep.

What a Tumor Marker Is

Cells make proteins. Some cancer cells make unusual amounts of particular proteins, and some of those leak into the blood where a lab can measure them. That is the whole mechanism — and every limitation follows from it.

Healthy tissue makes many of these same proteins in smaller amounts, so nobody's level is zero. Inflamed, irritated, or simply busy tissue can make more. A small early tumour may not make enough to move the number at all. The result is a test that produces a number with a reference range next to it, which looks far more decisive on a report than it deserves to.

The Five You'll See Most Often

Reference ranges are general laboratory conventions and vary between laboratories — your own report's range is the one that applies to you.

You may also see CA 15-3, associated with breast cancer. It is used mainly to track known disease rather than to look for new disease, and the same caveats in this article apply to it.

One quirk worth knowing about CA 19-9: roughly 5–10% of people do not produce the antigen at all, for genetic reasons. In those people the test reads low no matter what is happening in the pancreas. It is a test that can be structurally blind in a way you would never know from the report.

Why a Raised Marker Isn't a Diagnosis

Two words explain most of the confusion, and they are worth understanding in plain terms.

Sensitivity is how often a test flags the disease when the disease is present. Low sensitivity means the test misses real cases — a normal result that means nothing.

Specificity is how often a test stays quiet when the disease is absent. Low specificity means the test flags people who are fine — a raised result that means nothing.

Tumor markers are mediocre at both, and there is a third problem that matters even more: how rare the disease is. Ovarian cancer affects a very small share of women. If you test a large group of healthy women with a test that flags a few percent of healthy people, the great majority of the raised results will be women without ovarian cancer — even if the test is working exactly as designed. The maths is unforgiving, and it is why a test can be simultaneously well-made and useless as a screen.

What happens next is the actual harm. A raised marker in a healthy person leads to repeat blood draws, scans, sometimes a biopsy, sometimes surgery on an ovary that turned out to be fine — plus months of fear. Screening is only worth doing when finding something changes the outcome for the better, and often enough here, it doesn't.

The Harmless Things That Raise These Numbers

These are not rare exceptions. They are common:

• Smoking raises CEA. This is so consistent that laboratories apply a separate, higher reference range for smokers — around 5.0 ng/mL against 2.5 ng/mL for non-smokers. If your report shows one range only, ask which one was used.
• CA 125 rises with endometriosis, benign ovarian cysts, uterine fibroids, and pelvic inflammation. All are common, none are cancer. This is the single largest source of frightening false alarms in women's screening.
• CA 125 also shifts across the menstrual cycle, running higher during menstruation. In premenopausal women it is best drawn mid-cycle — roughly days 7–14, after your period ends. If you are booking a screening and this marker is in the panel, say so when you schedule; timing the draw is free and removes a real cause of false positives.
• PSA rises with benign prostatic enlargement, prostatitis, urinary infection, and after ejaculation or cycling. It is a prostate test, not a prostate cancer test.
• AFP rises in pregnancy and with hepatitis and other liver inflammation.
• CA 19-9 rises with bile duct obstruction — including from gallstones — and other benign gastrointestinal conditions.

What the Guideline Bodies Actually Say

This is where most articles wave at "guidelines" without naming one. The specifics matter, so here they are:

• The US Preventive Services Task Force recommends against screening for ovarian cancer in asymptomatic women — a Grade D recommendation, meaning it judges the harms to outweigh the benefits. CA 125 is one of the tests that recommendation covers (2018 statement).
• The National Cancer Institute states that tumor markers generally perform poorly as screening tests, because they are neither sensitive nor specific enough and are raised by benign conditions.
• The American Society of Clinical Oncology does not recommend CEA for colorectal cancer screening. ASCO supports CEA for monitoring patients with known colorectal cancer — a different job entirely.
• PSA is the partial exception. The USPSTF frames PSA screening as an individual decision for men aged 55–69, to be made with a clinician after weighing benefits and harms (Grade C), and recommends against it for men 70 and over (2018 statement, currently under review). That is the strongest position any of these markers holds, and it still stops short of "everyone should get it".

Notice the pattern. Not one of these bodies says the tests are worthless. They say the tests are wrong for finding cancer in people who have no reason to suspect it.

Where Tumor Markers Genuinely Work

Change the question and the same test becomes valuable — because the maths changes with it. In someone already diagnosed with cancer, the disease is not rare; it is certain. The number is no longer being asked "is there cancer?" but "is this cancer doing what we expect?"

• Monitoring known cancer. A marker falling after surgery or during chemotherapy is evidence the treatment is working. Rising says the opposite. This is a comparison against your own earlier number, not against a population range.
• Watching for recurrence. After successful treatment, a marker that starts climbing can signal return before symptoms appear. CEA in colorectal cancer follow-up is the standard example — the same test ASCO rejects for screening.
• Occasionally, high-risk surveillance. In specific groups — for example AFP alongside ultrasound in people with cirrhosis or chronic hepatitis B, a population well represented across Asia — markers have a defined role. That is surveillance of a known-risk group by a treating doctor, not general screening.

The thread through all three: the marker is being read as a trend in one person over time, not as a verdict from one draw.

So Why Do We Offer Them?

A fair question to ask of anyone selling a test their own cited sources decline to recommend. The honest answer has three parts.

First, a tumor marker never appears alone in our screening. The Tumor Markers Panel is a catalog item, and markers also sit inside broader blood work — our Comprehensive Blood Work Panel is $249 — where a single number is one weak signal among dozens, read next to liver function, inflammatory markers, and whatever the imaging in your package shows. A CA 19-9 that is mildly raised means something very different when the abdominal ultrasound is clean than when it is not.

Second, every package includes a physician consultation. A marker handed over as a PDF with a red flag next to it is a source of harm. The same marker explained by a doctor who can say "you smoke, this is your expected range, here is what we watch" is not. That conversation is the product; the number is an input to it.

Third, a first result becomes your baseline. Trends are where these tests work, and a trend has to start somewhere.

What we will not tell you is that a normal panel means you do not have cancer. It doesn't, and anyone who implies otherwise is selling you something. If you want tests with real evidence behind them for finding cancer early, they are unglamorous and specific: colonoscopy, mammography, low-dose lung CT for heavy smokers, HPV testing. Several are in our full-day packages — the guide to digestive tests and blood work for visitors covers how the endoscopy side fits together.

What to Do With This

If you are booking a screening in Taiwan and a tumor marker panel is in it, three things make the result more useful and less frightening:

• Tell them if you smoke, so CEA is read against the right reference range.
• If you are a premenopausal woman, ask to have CA 125 drawn mid-cycle — and if the appointment falls during your period, ask whether that marker can be timed differently.
• Bring your history. Endometriosis, fibroids, hepatitis B, gallstones, a benign prostate diagnosis — every one of them changes how a number should be read. A doctor who does not know cannot read it correctly.

And if a marker comes back raised: it is a reason for a conversation, not a conclusion. The most likely explanation, statistically, is one of the benign causes above. Reports are written in English and a coordinator confirms preparation and visit specifics when you book — including what is in each panel, which is worth asking about directly rather than assuming. You can see the current lineup on the packages page, and the travelers' guide to health screening in Taiwan covers the logistics of getting here.

This article is for general information and does not replace medical advice. Discuss your own screening needs and risk factors with a qualified clinician.