From the Journals

Study: Test carcinoembryonic antigen in colon cancer after surgery


Key clinical point: Measurement of CEA levels in patients with colon cancer appears to have value postoperatively but not preoperatively.

Major finding: Patients with abnormal postoperative CEA levels were less likely to reach 3-year recurrence-free survival than were those with normal levels (74.5% vs. 89.4%).

Study details: Retrospective, single-center study of 1,027 patients with stage I-III colon cancer (50.4% male; median age, 64 years).

Disclosures: The study authors report no relevant disclosures. The National Institutes of Health/National Cancer Institute funded the study in part.

Source: Konishi T et al. JAMA Oncol. 2017. doi: 10.1001/jamaoncol.2017.4420.

View on the News

CEA holds potential as colon cancer biomarker

Multiple organizations recommend that patients with nonmetastatic colorectal cancer undergo carcinoembryonic antigen (CEA) measurement prior to surgery, and there’s talk that it should be added to the staging system of the American Joint Committee on Cancer.

This new study, however, provides a challenge to the use of preoperative CEA tests in this population. As the researchers report, patients with normalized CEA levels postsurgery – even those with abnormal levels before surgery – were more likely to reach recurrence-free survival for 3 years.

The findings spotlight the value of real-world evidence and real-world data rather than statistics gathered during a controlled clinical trial.

There are limitations to this study, such as generalizability (it was performed at a single center), loss of patients to CEA follow-up postsurgery, and timing (the study period began a decade ago).

Still, the findings, once confirmed, could change clinical practice, and they point to the potential use of CEA as a biomarker early-warning sign that a tumor may recur.

Rebecca Anne Miksad, MD, MPH, is with Harvard Medical School in Boston (leave of absence from 2017 to 2018) and Neal J. Meropol, MD, is with Case Western Reserve University in Cleveland. They disclose employment by Flatiron Health. These comments are taken from their editorial accompanying the study by Konishi et al. in JAMA Oncology, doi:10.1001/jamaoncol.2017.4420.



A new study challenges the practice of measuring carcinoembryonic antigen (CEA) in patients with colon cancer prior to surgery, as is currently recommended by some guidelines. Researchers found that the levels do have predictive value about risk of recurrence when tested after surgery, but they challenge its use before.

“Consistent with the literature, our data show that postoperative CEA is more informative than preoperative CEA,” the study authors wrote. “Emphasis should be placed on postoperative CEA, and in the setting of modern high-quality imaging, we question the utility of measuring preoperative CEA.”

The study, which was published online in JAMA Oncology, noted that testing for CEA – a colorectal cancer tumor marker – had more value prior to the current era of imaging because it could indicate the need to widen a search for metastases.

“In the era of high-quality computed tomography (CT), the utility of measuring preoperative CEA is less obvious because an elevated preoperative CEA with a normal CT scan does not preclude surgery with curative intent,” wrote Tsuyoshi Konishi, MD, of Memorial Sloan-Kettering Cancer Center in New York and his coauthors.

CEA levels can return to normal after colon cancer surgery – but not always. According to the study, the National Comprehensive Cancer Network, the American Society of Clinical Oncology, and the European Group on Tumor Markers recommend its measurement prior to surgery in patients with nonmetastatic cancer.

For the new study, the authors sought to understand the predictive value of various CEA levels both before and after surgery.

The researchers retrospectively tracked 1,027 patients at a single center with stage I-III colon cancer (50.4% male; median age, 64 years). (Another 221 patients had been removed because of exclusion criteria, such as recent malignancy treatment and preoperative chemotherapy or radiation treatment.)

Nearly 70% of the 1,027 patients had normal preoperative CEA levels, and the rest had elevated levels.

Of the 312 patients with elevated preoperative CEA levels, 113 were lost to follow-up postoperative CEA tests. In the remaining patients, CEA levels returned to normal in 71%.

Patients were followed for a median of 38 months; 10.3% had recurrences, and 4.6% died. The 3-year recurrence-free survival rate was 88.4%.

Researchers found that those with normal preoperative CEA levels were more likely to reach recurrence-free survival at 3 years (89.7%) than were those who had elevated preoperative CEA levels (82.3%; P = .05). The higher level wasn’t much different than the recurrence-free survival level in those whose CEA levels had normalized after the operation (87.9%; P = .86).

Those with elevated CEA levels after surgery had a lower likelihood of reaching recurrence-free survival at 3 years (74.5%) than did those who had normal postoperative CEA levels (89.4%).

The study also linked abnormal postoperative CEA levels to shorter recurrence-free survival (hazard ratio, 2.0; 95% confidence interval, 1.1-3.5). Normalized postoperative CEA levels, however, were not linked to shorter recurrence-free survival (HR, 0.77; 95% CI, 0.45-1.30).

The researchers noted that they didn’t control for factors such as smoking (which increases CEA levels) and disease like gastritis and diabetes, which can produce misleading CEA levels. “Rather, this study is pragmatic,” the researchers wrote, “and represents what is likely to be seen in real-world colorectal cancer care.”

Moving forward, the researchers wrote, “patients with elevated postoperative CEA tend to experience recurrence early, which might justify a risk-adjusted and individualized surveillance strategy.”

The study authors report no relevant disclosures. The National Institutes of Health/National Cancer Institute funded the study in part.

SOURCE: Konishi T et al. JAMA Oncol. 2017. doi: 10.1001/jamaoncol.2017.4420.

Next Article: