Do CEA Levels Fluctuate During Chemo?

CEA levels can and do fluctuate during chemotherapy, sometimes dramatically. A well-documented phenomenon called a “CEA surge” causes levels to spike early in treatment even when the cancer is actually responding, with median increases measured in the hundreds of percent in some patients. The pattern catches many people off guard, because a rising tumor marker during treatment feels like the worst possible news. Understanding why it happens and how oncologists interpret it can spare you weeks of unnecessary dread.

What a CEA Surge Looks Like

CEA, or carcinoembryonic antigen, is one of the most commonly used blood markers for monitoring colorectal cancer during and after treatment.¹ The general expectation is straightforward: if treatment is working, CEA should go down. If the cancer is growing, CEA should go up. In practice, it is messier than that.

In a study of patients receiving oxaliplatin-based chemotherapy for metastatic colorectal cancer, about 15% experienced a transient spike in CEA despite having an objective treatment response on imaging. The median rise from baseline was 263%, with a range from 24% to 632%. These peaks occurred anywhere from 13 to 56 days after starting treatment.² A separate study defined the surge more precisely as a rise of more than 20% from baseline followed by a drop of more than 20% below baseline in subsequent readings. In that cohort, 10 out of 89 evaluable patients showed the pattern, and the surge resolved within four months in every case. All 10 patients were getting clinical benefit from their chemotherapy.³

The critical takeaway is that a rising CEA number in the first weeks of chemotherapy does not automatically mean the disease is progressing. It can be the opposite: a marker of the cancer responding and releasing its stored antigen into the bloodstream as cells die. The timing matters. A surge that peaks within the first couple of months and then falls is a fundamentally different signal from a CEA that steadily climbs over many months.

Telling a Surge Apart from Real Progression

This is the question that actually keeps patients and their oncologists alert: when a CEA reading goes up during treatment, how do you know whether it is a benign surge or a sign that the cancer is winning? The answer involves both the magnitude and the trajectory of the change, and it depends on what treatment regimen is being used.

A large analysis of metastatic colorectal cancer patients looked at CEA changes from baseline to the first restaging scan. For patients on chemotherapy alone, those whose cancer was not progressing saw a median CEA decrease of about 53%, while patients whose disease was progressing had a median increase of roughly 24%. The researchers identified a threshold of about a 7.5% decrease in CEA as the best cutoff for distinguishing the two groups. Patients whose CEA had not fallen by at least that amount were substantially more likely to have progressive disease on their scans.⁴

Interestingly, the picture changes when anti-VEGF therapy (a targeted drug like bevacizumab) is added. In those patients, even the ones with progressive disease showed a median CEA decrease of about 45%, and the non-progressing group showed a much steeper drop of around 72%. The useful cutoff shifted to a 62% decrease, meaning a smaller CEA drop could actually signal progression when a targeted agent is involved.⁴ This has practical consequences: you cannot apply one universal rule for what a “good” CEA trajectory looks like. The treatment regimen changes the goalposts.

A separate study examining patients who received neoadjuvant chemotherapy (sometimes including bevacizumab) before liver surgery for colorectal metastases found that a CEA decrease of 50% or more from baseline was a meaningful dividing line for predicting longer survival.⁵ These numbers converge on a rough principle: in most contexts, oncologists expect a responding patient’s CEA to eventually drop by at least half. But that eventual drop may follow an early spike, and the timing of each blood draw relative to chemotherapy cycles affects the number you see.

Why the Numbers Jump Around Even When Nothing Has Changed

Beyond the tumor itself, several non-cancer factors push CEA levels up or down in ways that have nothing to do with whether treatment is working. This is part of what makes CEA an imperfect marker, and it’s why oncologists treat it as one piece of a puzzle rather than a standalone verdict.

Smoking is one of the most consistent confounders. Smokers tend to carry higher baseline CEA levels than nonsmokers, and this effect does not disappear during treatment. If you quit smoking after starting chemo, your CEA might fall partly because of the treatment and partly because you stopped smoking. If you relapse to smoking, it may bump your number up by an amount that has nothing to do with the cancer. Liver disease from any cause, including alcohol use, can also raise CEA. Active inflammatory bowel conditions like ulcerative colitis tend to push CEA higher during flares, and levels typically settle down with remission. Even pancreatitis and certain benign colonic polyps can cause mild elevations.⁶

Chemotherapy itself can temporarily impair liver function, and the liver is responsible for clearing CEA from the blood. When liver enzymes are elevated due to drug toxicity, you might see a CEA bump that reflects slower clearance rather than increased production. Dehydration, which is common during treatment, can also concentrate blood proteins and produce slightly higher readings. None of these factors mean the cancer is growing, but they all contribute to the frustrating reality that any single CEA reading can be noisy.

Lab-to-Lab Variation Is Larger Than Most People Realize

There is another source of fluctuation that has nothing to do with your body at all: the assay itself. Different laboratories use different test kits to measure CEA, and these kits do not always agree. A multicenter study that compared CEA results across labs found a coefficient of variation of over 60% between laboratories when each used its own local standard. That means the same blood sample could produce readings that varied wildly depending on which lab processed it. When all labs converted to a shared international reference standard, the variation dropped to about 22%, which is a major improvement but still substantial.⁷

The practical implication is simple but important: always try to get your CEA drawn at the same lab, using the same assay, throughout treatment. Switching labs between readings introduces variation that can make a stable number look like it jumped, or mask a real change. If you do switch labs for any reason, mention it to your oncologist so the comparison between readings is interpreted with appropriate caution.

How Often CEA Should Be Checked During Treatment

There is no single universal schedule for CEA monitoring during chemotherapy, and the frequency depends on the clinical setting. During active treatment for metastatic disease, many oncologists check CEA before each cycle or at least monthly, because the trend line matters more than any individual reading. For surveillance after surgery without ongoing chemo, intervals stretch out, with major guidelines recommending checks every few months for the first several years.

The key concept is that a trend over three or more measurements is more informative than a single value. One elevated reading triggers attention. Two elevated readings in a row raise concern. A sustained, steadily climbing pattern over several months is what genuinely signals progression. Oncologists rarely change treatment based on a single CEA result alone; imaging with CT or PET scans remains the gold standard for confirming what the blood marker suggests. That said, a persistently rising CEA is often the first clue that prompts a scan earlier than originally scheduled.

The Emotional Toll of a Rising Number

The anxiety that a rising CEA produces is not a trivial side effect of monitoring. A case report documented the psychological cascade that can follow an unexpectedly elevated CEA level: anxiety, a rush of further testing, specialist referrals, and weeks of waiting, all triggered by a number that ultimately had a benign explanation.⁸ During chemotherapy, this cycle can repeat every time blood is drawn. A patient who understands that early surges are possible and that single-point readings are unreliable is better equipped to manage that waiting period without spiraling.

If your oncologist seems less alarmed about a CEA bump than you feel, that calm is usually informed by the trajectory of your previous readings, your imaging results, and knowledge of whether the timing fits a surge pattern. It is entirely reasonable to ask: “Is this within the range of a treatment surge, or should we be concerned?” Getting that direct answer can short-circuit weeks of anxious guessing.

When CEA Is Not the Right Marker at All

Roughly 20% to 30% of colorectal cancers do not produce elevated CEA in the first place. If your CEA was normal before treatment started, a normal reading during chemo tells you very little about whether the cancer is responding. Oncologists rely on imaging in those cases. Similarly, some cancer types do not express CEA at clinically useful levels, and using it as a monitoring tool for those cancers adds confusion without adding information.

Even among cancers that do produce CEA, the marker reflects total tumor burden imperfectly. A large liver metastasis may produce a lot of CEA while a peritoneal deposit of the same size may not, because of differences in how CEA enters the bloodstream from different sites. This contributes to cases where imaging and CEA seem to tell different stories. One study found that the correlation between CEA levels and tumor size was reasonably strong in patients with liver metastases but weak in patients whose disease was limited to sites other than the liver.⁹ Where the cancer is growing matters for how accurately CEA reflects what is happening.

Circulating Tumor DNA as a Complementary Marker

Circulating tumor DNA, or ctDNA, is increasingly being measured alongside CEA. It detects fragments of DNA shed by tumor cells into the bloodstream, and in theory it responds more quickly and specifically to changes in tumor burden. A study comparing the two markers in metastatic colorectal cancer found an overall concordance rate of about 76%, meaning they agreed roughly three-quarters of the time.⁹ That sounds decent, but the quarter of the time they disagreed is where things get clinically interesting.

The disagreement was not random. In patients with liver metastases, the correlation between the two markers was stronger. In patients without liver or lymph node metastases, ctDNA and CEA often told different stories, with CEA maintaining a decent correlation to tumor size in that subgroup while ctDNA did not. The implication is that neither marker is universally superior; they measure different signals. ctDNA may eventually prove more useful for detecting minimal residual disease after surgery, while CEA remains valuable for tracking bulky liver-dominant disease. During chemotherapy, a rising CEA paired with a falling ctDNA, or vice versa, adds complexity that oncologists are still learning to interpret.

Access to ctDNA testing is expanding but remains inconsistent. Not every cancer center offers routine ctDNA monitoring, and insurance coverage varies. For most patients on standard chemotherapy regimens, CEA remains the primary circulating biomarker, with all of its imperfections. But the field is moving toward multimarker panels that combine CEA, ctDNA, and imaging to give a more complete picture of treatment response. That shift may eventually reduce the anxiety surrounding any single CEA reading, because no one number will carry as much weight on its own.

How Liver Metastases Change the CEA Picture

The liver deserves special mention because it plays a dual role in CEA dynamics during chemotherapy. On one side, the liver is the primary organ that clears CEA from circulation. Chemotherapy drugs, particularly oxaliplatin-based regimens, can cause liver inflammation or fatty changes that impair this clearance, temporarily raising CEA levels even if tumor burden is stable or shrinking. On the other side, colorectal cancer metastasizes to the liver more than to any other organ, and liver metastases tend to be heavy CEA producers.

This creates a situation where patients with colorectal liver metastases may see especially volatile CEA readings during treatment: the tumors are producing CEA, the chemotherapy is killing some of those cells and releasing more CEA, and the liver’s ability to clear the antigen is itself compromised by the drugs. Early detection of new liver metastases using CEA alone remains difficult, even after decades of use.¹⁰ Imaging remains essential for catching new liver lesions, and a stable or slowly rising CEA in the presence of liver disease should prompt imaging rather than reassurance.

For patients undergoing neoadjuvant chemotherapy before planned liver surgery, the degree of CEA decline carries prognostic weight. A drop of 50% or more from baseline has been linked to better overall survival after the liver resection, suggesting that the magnitude of CEA response partly reflects the biology of how the tumor responds to treatment.⁵ Patients whose CEA does not fall much despite completing a full course of chemo may have less favorable biology, though individual exceptions exist and CEA is never the sole factor in surgical decision-making.