Cancer Antigen 15-3 Test · Blood

If you have been diagnosed with breast cancer, or you are being monitored after treatment, CA 15-3 (Cancer Antigen 15-3) gives you a number that reflects how much tumor activity is happening in your body right now. A rising number often means the disease is growing. A falling number often means treatment is working. That direct, serial feedback is something imaging alone cannot always provide in real time.

What CA 15-3 cannot do is screen for breast cancer in someone who has never been diagnosed. Its sensitivity for early, localized tumors is low, around 31% in one large study of newly diagnosed patients. That means roughly seven out of ten early breast cancers produce a normal CA 15-3 reading. The test earns its value after diagnosis, not before it.

What CA 15-3 Actually Measures

CA 15-3 is a piece of a large protein called MUC1 (mucin 1), a glycoprotein that normally sits on the surface of cells lining your breast ducts, lungs, and digestive tract. In healthy tissue, MUC1 plays a protective role. When breast cancer develops, the tumor cells overexpress MUC1 and shed fragments of it into the bloodstream. The more tumor cells you have, the more MUC1 fragments circulate, and the higher your CA 15-3 number climbs.

Because CA 15-3 reflects the total amount of MUC1 being shed, it is best understood as a surrogate for tumor burden, the total volume of active cancer in your body. It does not measure inflammation, hormone levels, or organ function the way most blood tests do. It measures how much of a specific cancer-related protein is floating in your blood.

Prognosis at Diagnosis

When CA 15-3 is measured before surgery in someone with newly diagnosed breast cancer, a high reading carries real prognostic weight. In a study of 600 patients followed for a median of about six years, those with preoperative CA 15-3 above 30 U/L had roughly 2.4 times the risk of death compared to those below that threshold, even among women whose cancer had not yet spread to lymph nodes. That association held after adjusting for tumor size, nodal status, and age.

A separate study of 470 Chinese breast cancer patients found that elevated preoperative CA 15-3 roughly doubled the risk of recurrence (about twice the hazard) and the risk of death (about twice the hazard), independently of lymph node involvement and tumor grade. In the largest dataset available, the Korean Breast Cancer Society Registry covering over 149,000 surgically treated patients, elevated preoperative CA 15-3 was a significant prognostic factor in hormone receptor positive (luminal) subtypes but not in triple negative breast cancer.

If you are facing a new breast cancer diagnosis, knowing your preoperative CA 15-3 adds a layer of risk information that tumor size and lymph node status alone do not fully capture. A high reading does not change your diagnosis, but it may influence how aggressively your oncologist approaches treatment planning.

Monitoring Recurrence and Treatment Response

The strongest case for CA 15-3 testing is in serial monitoring. In a Belgian cohort of 730 early breast cancer patients followed after surgery, 37% of metastatic relapses were first detected because of a rising CA 15-3, before symptoms or scheduled imaging revealed the problem. Patients whose metastatic disease was caught early by marker surveillance had a median overall survival of 35 months, compared to 22 months when metastases were found by other means.

The rate of change matters even more than the absolute number. In a study of 67 patients with confirmed recurrence matched against controls, a CA 15-3 velocity (rate of rise) above 2.5 U/mL per year, combined with a similar velocity measure for CEA (carcinoembryonic antigen, another tumor marker), predicted recurrence with 94% sensitivity and 73% specificity. The area under the curve for this velocity approach was 0.85, meaning it correctly classified the vast majority of patients.

In metastatic breast cancer, serial CA 15-3 tracks treatment response in real time. In a study of 53 women with metastatic disease, a 25% or greater decrease in CA 15-3 closely paralleled radiologic evidence of tumor shrinkage, while a 25% or greater increase tracked disease progression. CA 15-3 outperformed CEA in correlating with clinical course in that setting.

What CA 15-3 Cannot Tell You

CA 15-3 is not specific to breast cancer. Elevated levels appear in ovarian, lung, liver, and gastrointestinal cancers. In a study of 500 patients with non-breast malignancies, CA 15-3 was elevated across virtually all epithelial tumor types, particularly when metastases were present. Benign conditions can also raise the number: chronic liver disease, tuberculosis, sarcoidosis, lupus, and benign breast disease have all been associated with mild elevations, with roughly 3% of benign cases exceeding 40 U/mL in one large series.

Equally important, a normal CA 15-3 does not rule out cancer or recurrence. In a PET/CT correlation study of 154 patients with recurrent breast cancer, more than half of those with proven recurrence on imaging had normal CA 15-3 levels. The test completely misses many locoregional and lung-only recurrences. A normal result is not a clean bill of health.

CA 15-3 vs. CA 27.29

CA 27.29 targets the same MUC1 protein as CA 15-3, using a slightly different antibody. In a massive concordance analysis of 37,652 paired measurements, the two tests were highly correlated (a near-perfect statistical correlation of 0.967), with 93.7% of paired results falling in the same reference range category. Trends over time were similarly concordant: when one marker doubled or halved, the other almost always did the same. Ordering both tests routinely adds no meaningful clinical information.

Reference Ranges

CA 15-3 reference ranges vary by assay manufacturer, and no universal, demographic-stratified intervals exist. The values below represent commonly used clinical thresholds drawn from published studies, not a single guideline. Your lab may report slightly different numbers depending on the assay platform used.

These thresholds are orientation, not diagnosis. Compare your results within the same lab over time for the most reliable trend. A single reading near a cutpoint tells you far less than two or three readings spaced weeks apart.

When Results Can Be Misleading

The within-person biological variation for CA 15-3 is about 6.2%, but when you add in the analytical imprecision of the assay (about 11.2%), two consecutive results need to differ by roughly 30% before you can be statistically confident the change is real. Small fluctuations of a few units are noise, not signal.

• Liver disease: Chronic hepatitis and cirrhosis elevate CA 15-3 in up to 13 to 43% of patients, mimicking tumor-related elevation.
• Other cancers: Ovarian, lung, and gastrointestinal cancers can all push CA 15-3 well above normal, so an elevated result does not automatically point to the breast.
• Pregnancy and lactation: Some CA 15-3 assay platforms show false elevations during pregnancy and breastfeeding due to antibody cross-reactivity. This is assay-dependent, so check which platform your lab uses.
• Metformin: A large randomized trial showed that metformin modestly lowered CA 15-3 by about 6% over six months, independent of tumor characteristics. If you take metformin, this small downward shift could theoretically mask a small true rise.

Age, BMI, fasting state, time of day, and common cardiovascular or metabolic conditions do not appear to meaningfully affect CA 15-3 levels based on current evidence.

Tracking Your Trend

A single CA 15-3 value is a snapshot. Its real power comes from tracking it over time. Because the index of individuality for this marker is low (around 0.2), population-based cutoffs are less informative than your own personal baseline and trajectory. Two people with identical readings may have very different clinical stories; what matters is how your number moves.

If you are being monitored after breast cancer treatment, a reasonable cadence is every three to six months during the first two to three years after treatment, when recurrence risk is highest, then every six to twelve months thereafter. Always use the same lab and the same assay platform so your numbers are directly comparable. A sustained rise of 25% or more across two or more consecutive draws is the threshold most clinicians use to trigger further investigation.

What to Do With an Abnormal Result

A single mildly elevated CA 15-3 (say, 32 U/mL when your lab's upper limit is 30) should prompt a repeat draw in four to six weeks, not panic. If the repeat confirms the elevation or shows a rising trend, the next step is imaging, typically a CT scan, bone scan, or PET/CT depending on your clinical history. Your oncologist may also order CEA alongside CA 15-3, since combining the two markers improves detection of recurrence.

If your CA 15-3 is markedly elevated (above 50 U/mL) and you have a history of breast cancer, imaging and an oncology consultation should happen promptly. If you have no history of breast cancer but find an elevated CA 15-3, remember that many non-breast conditions can cause this. A thorough workup including liver function tests, imaging of the chest and abdomen, and a gynecologic evaluation is more appropriate than assuming breast cancer.

For anyone tracking this marker, the combination of CA 15-3 trend, CEA trend, imaging, and clinical symptoms together form the picture. No single data point drives a decision.