What Is the CEA Test? In Which Cancers Is an Elevated Level Important?
In this article
What is the CEA test, and what is it used for?
CEA is not a test that diagnoses cancer—and this is the most important sentence in this article. CEA is a monitoring tool: its value comes not from a single number but from the trend over time in the same patient, at the same laboratory. Not understanding this distinction causes harm in both directions: unnecessary panic in healthy people, and postponed colonoscopies in cancer patients who say, "My level is normal."
- CEA is not a screening test. In stage 1 colorectal cancer, only 4% of cases have elevated CEA (at a threshold of 5 ng/mL). In other words, it misses almost all early cancers.
- The normal level depends on whether you smoke: in nonsmokers ≤3 ng/mL; it may be somewhat higher in smokers, generally below 5 ng/mL.
- Use the same laboratory. The CEA result varies with the method used; values from different laboratories cannot be compared directly.
- CEA falls rapidly after surgery. Its half-life is 3-13 days; it is expected to return to normal within approximately 6 weeks. If it does not, residual disease may remain.
- Guidelines recommend CEA every 3-6 months for 5 years in stage 2 and 3 colorectal cancer —there are insufficient data for stage 1 and for patients whose liver metastases have been removed.
- And, to be honest: intensive follow-up detects more recurrences that can be operated on, but large randomized studies have not demonstrated a survival benefit.
- A point that needs correcting: In hereditary colon cancer syndromes such as Lynch syndrome, CEA is not a screening tool. Screening in these patients is colonoscopy.
What is CEA?
Carcinoembryonic antigen is a glycoprotein involved in cells sticking to one another. The meaning of its name also explains its biology: it is normally produced in a baby's digestive tract tissue in the womb, and production stops before birth. It is therefore present at very low levels in the blood of healthy adults. Some cancer cells restart this production, releasing the molecule into the bloodstream.
For a test to be used in diagnosis, it must do two things: in people with the disease it must be elevated (sensitivity), and in those without the disease it must remain normal (specificity). CEA is inadequate in both respects.
The sensitivity problem: CEA does not rise in most early-stage cancers.
The specificity problem: CEA may rise in many noncancerous conditions—smoking, chronic lung disease, cirrhosis, inflammatory bowel disease, pancreatitis, ulcers, low thyroid hormone and kidney failure.
Combining these two problems means: in a healthy person, CEA measured "as a checkup" has a high likelihood of missing cancer and producing a false alarm.
Why can it not be used for screening? The numbers
Turkish sources circulate the statement that "it makes sense to use CEA as a screening tool in people with hereditary colon cancer syndromes (such as Lynch syndrome)." This is not correct and may be harmful.
The recommended screening in Lynch syndrome is colonoscopy every 1-2 years from age 20-25; additional screening of the relevant organs is added to this. No blood tumor marker is included in accepted follow-up protocols.
Why does this matter? As the chart above shows, CEA misses early cancers. Postponing a due colonoscopy because "my CEA was normal" is the most dangerous mistake discussed in this article.
What is the normal level? And why is a single number insufficient?
| Situation | Expected CEA level |
|---|---|
| Healthy nonsmoker | ≤3.0 ng/mL —97% of healthy people have values below this |
| Smoker | May be somewhat elevated, generally below 5.0 ng/mL |
| Benign conditions | May rise but rarely exceeds 10 ng/mL |
| Above 20 ng/mL | In a patient with compatible symptoms, strongly suggests the presence of cancer; often indicates disease that has spread |
1. The same laboratory. In the reference laboratory's own words, CEA values are "method-dependent; therefore, the same method should be used for serial monitoring of patients." Comparing two CEA values measured at different hospitals is like measuring with two different rulers. Patients under follow-up should use the same laboratory whenever possible.
2. The trend, not a single value. The same source states: "single values are less informative than changes assessed over time." A single result of 4.2 ng/mL says almost nothing; a series of three measurements going from 2.1 → 3.8 → 6.5 says a great deal.
What should happen after surgery?
CEA's half-life is 3-13 days (the higher the preoperative value, the longer it takes to fall). The practical consequence of this kinetic behavior is that, if the cancer has been completely removed, CEA should return to normal within approximately 6 weeks.
What does it mean if it does not? Residual disease may remain. This finding requires additional imaging and assessment.
There are two reasons for measuring CEA before surgery. First, it establishes a baseline value for subsequent follow-up—because monitoring CEA would not be meaningful in a tumor in which it had never risen. Second, it provides prognostic information in its own right: in stage 3 patients, when preoperative CEA was above 5 ng/mL, the median time to recurrence was reported as 13 months, whereas when it was below 5, it was 28 months as reported.
How often should it be measured during follow-up? What the guidelines say
| ASCO | ESMO | |
|---|---|---|
| For whom? | Stage 2 and 3 colorectal cancer. "Insufficient data" are stated for stage 1 and resected metastatic disease. | Surgically resected colorectal cancer |
| CEA frequency | Every 3-6 months for 5 years (with medical history and examination) | Every 3-6 months for 3 years; then every 6-12 months in years 4 and 5 |
| Imaging | Chest and abdominal CT: Once a year for 3 years; every 6-12 months in high-risk patients | CT: in high-risk stages every 6-12 months for 3 years |
| Colonoscopy | In year 1, then every 5 years | — |
1. CEA monitoring is not for everyone. In stage 1 colorectal cancer and resected metastatic disease, guidelines state that there are insufficient data to recommend CEA monitoring. "I have had cancer, so I should have CEA checked" is not automatically correct.
2. Beyond five years is not defined. The guidelines' schedule is limited to 5 years; beyond that, decisions are made individually.
Does intensive follow-up really help? An honest answer
This section is almost never included in patient information, yet is the most honest part. After colorectal cancer surgery, intensive versus less frequent follow-up was tested in randomized studies, with a counterintuitive result.
| Study | What was done? | Result |
|---|---|---|
| FACS (JAMA 2014) |
1,202 patients, 4 groups: minimal follow-up / CEA alone / CT alone / CEA+CT | Rate of surgery for recurrence with curative intent: minimal follow-up 2.3%; CEA 6.7%; CT 8.0%; CEA+CT 6.6%. Mortality: 18.2% (intensive) versus 15.9% (minimal)—no difference. And the study's own conclusion: Combining CEA and CT offers no additional benefit. |
| COLOFOL (JAMA 2018) |
2,456 stage 2-3 patients; CT+CEA at months 6/12/18/24/36, or only at months 12 and 36? | No reduction in 5-year mortality was achieved. (We could not access the exact percentages, so we do not provide figures.) |
| Meta-analysis (2019, ~17 randomized studies, ~8,000 patients) |
Pooled comparison of intensive and less frequent follow-up | Overall survival HR 0.84 (0.74-0.96)—borderline favorable. However, cancer-specific survival HR 0.90 (0.77-1.04)—not significant. Curative reoperation 11.9% versus 6.2% (OR 2.23). |
Does intensive follow-up work? Partly. It detects more recurrences that can be operated on —a consistent and repeated finding. But large modern studies have not shown that this translates into survival.
Why might this be? A possible explanation is that some recurrences detected early are already treatable regardless of when they are found, while others cannot be cured no matter how early they are detected. Early diagnosis brings detection forward but does not always change the outcome.
So should there be no follow-up? No—and that is not what the data say. The schedule recommended by guidelines remains valid. The practical meaning of these findings is: testing more often than the guideline recommends provides no additional benefit. Having a CT scan every three months does not make you live longer than having one every six months—but produces more radiation, more cost and more anxiety.
CEA has risen—what happens now?
- First, confirm it. A single elevated value does not prompt a comprehensive investigation; the measurement is repeated at least two weeks later. Isolated rises are common and most resolve on their own.
- Look at the trend. What matters is not the absolute value, but a consistent rise across consecutive measurements. The threshold used in the FACS study was a rise of 7 units above baseline.
- Exclude confounding causes. Smoking, active infection, liver disease, a flare of inflammatory bowel disease, kidney failure and—as discussed below—a temporary chemotherapy-related rise.
- Perform imaging. Usually chest and abdominal CT.
- If imaging is clear but CEA continues to rise, PET-CT is considered. We could not find a verified figure for this step's detection rate, so we do not give a percentage.
In the first weeks after chemotherapy begins, CEA may rise temporarily without this meaning the disease is progressing—it is thought to be related to the breakdown of tumor cells.
Its frequency has been reported: at a threshold of 5 ng/mL, the false-positive rate is 16%, and at a threshold of 10 ng/mL, 4%. These rises resolve on their own. In contrast, the source states: "Rises above 15 ng/mL are unlikely to reflect anything other than recurrence."
The practical conclusion: in a patient who has just started chemotherapy, a CEA rise at the first checkup is not sufficient on its own to justify changing treatment. The measurement is repeated and assessed together with imaging.
In which cancers is it used, and in which is it not?
| Cancer | CEA's role |
|---|---|
| Colorectal cancer | The main area of use. Preoperative prognosis, postoperative assessment and monitoring for recurrence during follow-up. |
| Medullary thyroid cancer | It is a secondary marker—the main marker is calcitonin. Nevertheless, it provides information: CEA above 30 µg/L suggests spread beyond the thyroid, while above 100 µg/L suggests spread to lymph nodes or distant sites. "Doubling time" is also prognostic: the course of disease is substantially better in those with a calcitonin doubling time longer than 2 years than in those with a time shorter than 6 months. |
| Breast cancer | Routine measurement of tumor markers during follow-up is not recommended. In disease that has spread, it may be used with other markers to monitor treatment response. |
| Lung cancer | It is not a diagnostic or follow-up tool; it provides prognostic information (high values are associated with advanced stage, lymph node involvement and poorer survival). |
| Pancreatic cancer | CEA is not the marker for this cancer—the marker used is CA 19-9. |
| In body fluids | Measuring CEA in abdominal fluid (ascites) may help distinguish cancer-related fluid: sensitivity 51%, specificity 97% —an elevated value is meaningful, while a normal result does not exclude cancer. |
Frequently asked questions
- My CEA was 4,5. Do I have cancer?
- Most likely not. The 3-5 ng/mL range is common, especially in smokers, and many benign conditions (chronic lung disease, liver disease, inflammatory bowel disease, pancreatitis, ulcers, low thyroid hormone and kidney failure) may raise it to this level. Benign causes rarely raise CEA above 10 ng/mL. What you should do is not panic, but mention if you smoke, repeat the measurement at the same laboratory and assess it with your physician.
- I am healthy. Should I have CEA measured as a checkup?
- No—and not because the test is "expensive," but because it does not work for this purpose. Only 4% of stage 1 colorectal cancers have elevated CEA, so it misses the cancers for which detection matters most. In contrast, it frequently produces false alarms in healthy people, leading to unnecessary chains of investigations. Colorectal cancer screening uses fecal occult blood testing and colonoscopy, not CEA.
- Hereditary colon cancer runs in my family. Can I be monitored with CEA?
- No. In hereditary syndromes such as Lynch syndrome, accepted follow-up is colonoscopy every 1-2 years from age 20-25; screening for other relevant organs is added to this. No blood tumor marker is included in these protocols. A normal CEA result should never be a reason to postpone a colonoscopy that is due.
- I had surgery, but my CEA is still elevated. What does that mean?
- CEA's half-life is 3-13 days, and if the tumor has been completely removed, it is expected to return to normal within approximately 6 weeks. If it does not, residual disease may remain, requiring additional imaging. However, a value measured in the first days immediately after surgery may not yet have fallen—timing matters.
- My CEA is rising, but my CT scan is clear. What is done?
- This is common and generally managed step by step: first, confirm the measurement by repeating it at least two weeks later, exclude confounding causes (smoking, infection, liver or kidney problems, newly started chemotherapy), and if the rise continues, PET-CT is considered. In some patients, the cause is never found, and close monitoring continues.
- I started chemotherapy, and my CEA rose. Is treatment not working?
- Not necessarily. In the first weeks of chemotherapy, a temporary rise in CEA may occur, attributed to the breakdown of tumor cells. This false positivity occurs at a rate of 16% at a threshold of 5 ng/mL and 4% at a threshold of 10 ng/mL, and resolves on its own. In contrast, rises above 15 ng/mL are not expected to be due to a cause other than recurrence. Treatment is not changed on the basis of a single measurement.
- Will my CEA fall if I stop smoking?
- Somewhat elevated CEA (generally below 5 ng/mL) is expected in smokers. Levels are expected to decline after smoking cessation. Beyond this, your physician needs to know you smoke to interpret the result correctly —without that information, an innocent elevation may lead to unnecessary investigations.
- Why am I always told to use the same laboratory?
- Because CEA results depend on the measurement method, and values from different laboratories cannot be compared directly. Reference laboratories give this same warning: serial monitoring should use the same method. In follow-up, the main information comes not from a single number but from the trend over time—and if you measure with different rulers, that trend becomes unreliable.
Glossary
- CEA
- Carcinoembryonic antigen. A protein normally produced in the womb, production of which is restarted by some cancer cells.
- Tumor marker
- A substance measurable in blood that may be associated with cancer. It does not establish a diagnosis; it is used for monitoring.
- Sensitivity
- A test's rate of identifying people who have the disease. CEA has low sensitivity in early-stage disease.
- Specificity
- A test's rate of remaining normal in people without the disease. CEA has limited specificity.
- False positive
- An elevated test result despite the absence of disease.
- Half-life
- The time it takes for the amount of a substance in blood to fall by half. For CEA, this is 3-13 days.
- Curative surgery
- Surgery intended to eliminate the disease completely.
- Doubling time
- The time it takes for a marker to double. It has prognostic value in medullary thyroid cancer.
- HR (hazard ratio)
- A comparison of risk between two groups. If the confidence interval includes 1, the result is not statistically significant.
- PET-CT
- An imaging method showing sugar use, combined with CT. It may be used when CEA is rising while CT is clear.
DROZDOGAN Academy Commentary
CEA is one of the most frequently requested and misused tests in oncology. The problem is not the test itself but what is expected of it: it is expected to diagnose, whereas it can only monitor.
⚡ Why does this subject matter?
In Turkey, requesting tumor markers "as a checkup" is a common habit. It has two costs. The first is a false alarm: a CEA result of 4.8 in a healthy person may trigger weeks of investigations and anxiety. The second, more dangerous cost is false reassurance: normal CEA may cause a due colonoscopy to be postponed. Relying on early-stage sensitivity of 4% means missing precisely the cancer that should be detected.
✅ Strengths / positive developments
When used in the right patient and to answer the right question, CEA really is useful. Measured before surgery, it provides both a baseline for follow-up and prognostic information. Its postoperative kinetics are clear: its half-life is measured in days, it is expected to normalize in six weeks, and failure to do so is a concrete finding. The guideline schedule (every 3-6 months, stage 2-3, five years) is clear and practical. The prognostic value of doubling time in medullary thyroid cancer is a classic example of looking at a trend rather than a single number.
⚠️ Limitations / things not to overstate
Even one-third of stage 4 patients have normal CEA—so "my level is normal, everything is fine" can be said even in disease that has spread. There are no solid figures on how and how much benign causes raise it; we have only qualitative statements such as "rarely exceeds 10." And the limitation that deserves the most discussion is: intensive follow-up finds more recurrences that can be operated on, but large randomized studies have not demonstrated a survival benefit. In FACS, mortality was 18.2% versus 15.9%—not even lower in the intensive follow-up group. This is not a reason to abandon follow-up; it is evidence that there is no reason to exceed the guideline.
🩺 Our position in practice
- We do not request CEA "as a checkup" in healthy people. Neither to look for cancer nor for reassurance.
- We always measure it before surgery —for both a baseline and prognostic information. Monitoring CEA would not be meaningful in a tumor in which it had never risen.
- We ask patients to use the same laboratory and explain why.
- We do not start a chain of investigations on the basis of a single elevated value. First repeat the test, then exclude confounding causes.
- We do not interpret the first rise in a patient newly starting chemotherapy as progression.
- We do not request testing more often than guidelines recommend —there is no evidence that more frequent testing is better, but there is evidence that it produces more radiation and anxiety.
- In hereditary syndromes, we never replace colonoscopy with a blood test.
❓ Open questions
Studies unable to demonstrate a survival benefit from intensive follow-up cannot distinguish whether follow-up itself or our available tools are inadequate—new approaches such as circulating tumor DNA seem likely to reopen this question. There is still no proper table showing how much benign conditions raise CEA. And a Turkey-specific question remains: how many CEA tests are genuinely ordered for follow-up, and how many as a "checkup"? No study has measured this.
- Corrections: The statement "it makes sense to use CEA as a screening tool in hereditary colon cancer syndromes" was removed—accepted follow-up protocols do not include blood tumor markers; screening is colonoscopy. The reference range was completed (nonsmokers ≤3, smokers generally <5; method dependence and the same-laboratory rule). The time to normalization after surgery was corrected (half-life 3-13 days, expected normalization in approximately 6 weeks). Prostate, ovarian and liver cancer were removed from the list of cancers in which CEA is used; a clinical role for CEA could not be verified in these cancers. It was stated that the marker used in pancreatic cancer is CA 19-9.
- Additions: Stage-specific sensitivity (4% in stage 1)—the main reason why the test cannot be used for screening. ASCO and ESMO follow-up schedules and their applicability only to stage 2-3 disease. Randomized studies that could not demonstrate a survival benefit from intensive follow-up. A temporary chemotherapy-related rise. A stepwise approach to rising CEA. The prognostic value of preoperative CEA. Doubling time in medullary thyroid cancer. In headings, id tags were removed.
- Items for which no figures are given because they could not be verified: The exact percentages in the COLOFOL study (only the negative result was confirmed). Numerical results of the GILDA study. The hazard ratio from the Cochrane review. NCCN follow-up intervals (the text could not be accessed, so NCCN was not cited). The detection rate of PET-CT in patients with rising CEA but clear CT. How much each benign condition raises CEA. CEA performance in pleural fluid and cerebrospinal fluid.
- Turkey: No verifiable Turkey-specific data or guideline was found, so no numerical claims about Turkey were made. No information about SGK reimbursement has been provided.
- Internal links were partly retained from the existing article and are partly estimated; they should be checked before publication.
This article is for information and does not replace individualized medical advice. Your physician should interpret test results together with your clinical situation.
References
- Mayo Clinic Laboratories. Carcinoembryonic Antigen (CEA), Serum — test information and interpretation.
- Carcinoembryonic Antigen. StatPearls [Internet].
- Meyerhardt JA, Mangu PB, Flynn PJ, et al. Follow-up care, surveillance protocol, and secondary prevention measures for survivors of colorectal cancer: American Society of Clinical Oncology clinical practice guideline endorsement. J Clin Oncol.
- Argilés G, Tabernero J, Labianca R, et al. Localised colon cancer: ESMO Clinical Practice Guidelines for diagnosis, treatment and follow-up. Ann Oncol. 2020;31(10):1291-1305.
- Primrose JN, Perera R, Gray A, et al. Effect of 3 to 5 years of scheduled CEA and CT follow-up to detect recurrence of colorectal cancer: the FACS randomized clinical trial. JAMA. 2014;311(3):263-270.
- Wille-Jørgensen P, Syk I, Smedh K, et al. Effect of more vs less frequent follow-up testing on overall and colorectal cancer-specific mortality in patients with stage II or III colorectal cancer: the COLOFOL randomized clinical trial. JAMA. 2018;319(20):2095-2103.
- Intensive versus non-intensive follow-up after curative resection of colorectal cancer: a systematic review and meta-analysis. PLOS ONE. 2019.
- Wells SA Jr, Asa SL, Dralle H, et al. Revised American Thyroid Association guidelines for the management of medullary thyroid carcinoma. Thyroid.
- Idos G, Valle L. Lynch Syndrome. GeneReviews [Internet]. University of Washington, Seattle.
Sources
2. Carcinoembryonic antigen (CEA), serum. (n.d.)
mayomedicallaboratories.com/test-catalog/print/8521
3. Carcinoembryonic antigen. (2015, July 5)
muschealth.staywellsolutionsonline.com/Library/TestsProcedures/LabTests/167,carcinoembryonic_antigen
4. Carcinoembryonic antigen. (2016, September 8). Retrieved from
labtestsonline.org/understanding/analytes/cea/tab/test
Tumor markers. (2015, November 4)
5. cancer.gov/about-cancer/diagnosis-staging/diagnosis/tumor-markers-fact-sheet
Was this article helpful?
Your feedback helps us improve our content.
Reader comments
Share your comment
Your name and comment are published only after administrator approval. Your email address is never shown to visitors.
Be the first to comment.