Carcinoembryonic antigen (CEA) monitoring in patients with stage I-IV colorectal cancer has been, and remains, a controversial issue in oncology practice. Recommendations vary from bimonthly monitoring to no monitoring in the surveillance setting (for stage I-III disease). In the metastatic setting, there are no clear guidelines for CEA follow-up, although continued monitoring in such patients is common in the oncology community. This manuscript reviews the accuracy of CEA testing, its value as a prognostic indicator, and its role in surveillance and response assessment. The limitations of the test in the adjuvant and metastatic settings are illustrated through several case reports from the Colorectal Oncology Clinic at Roswell Park Cancer Institute. Guidelines for CEA monitoring are provided, based on a detailed literature review and institutional experience.
In their comprehensive review, Drs. Fakih and Padmanabhan consider a number of issues central to the putative utility of carcinoembryonic antigen (CEA) monitoring in caring for patients with colorectal cancer. They weigh the evidence for the use of CEA as a means of primary diagnosis for colorectal cancer, as a stage-independent prognostic marker, as a means of surveillance for recurrent disease after curative-intent surgery, and as a means to assess treatment response in advanced disease. In three of these four scenarios, the authors find the test lacking, and they give only a qualified endorsement for CEA when employed in serial testing as a surveillance tool. Does CEA testing still belong on our future appointment order sheets?
There is little argument about the uselessness of CEA as a screening tool. In early-stage disease the test lacks sensitivity. CEA levels can be elevated in diseases other than colorectal cancer as well as in "healthy" smokers. The authors catalog this information and appropriately dispense with the use of the test in individuals without a known diagnosis of colorectal cancer.
While there are numerous investigations in the literature examining the prognostic utility of CEA, the authors note that they yield conflicting results. In part, this is because even in the presence of advanced disease, at least 30% of patients have normal CEA levels.
The authors offer no definitive recommendation regarding how to factor CEA levels into a consultation at any stage of disease, and we agree that data regarding the usefulness of a single CEA level in this setting is rather inconclusive. We would add only that while CEA may offer some degree of stage-independent prognostic information, the strength of this information is insufficient to warrant practice-changing behavior in regard to the use of adjuvant chemotherapy when the test is done either before or after resection of either local/regional or metastatic disease.
Given the wide range of recurrence risk after surgery for colorectal cancer, there is a clear need for both additional prognostic markers and markers predictive of chemotherapy response and toxicity. Unfortunately, a solitary CEA level fails the test as an adequate prognostic or predictive measure. However, in patients with an established histologic diagnosis, repetitive testing may have some prognostic utility.
Drs. Fakih and Padmanabhan also recommend against the routine use of CEA to assess for disease response in advanced-stage colorectal cancer. They support this recommendation with evidence showing that CEA values are often discordant with radiographic and clinical response. Indeed, CEA levels may rise, fall, or remain the same in the face of symptomatic and radiographic improvement.
The authors appropriately stress the important point enumerated in the American Society of Clinical Oncology (ASCO) guidelines they cite that an individual should be considered to have progressive disease based on an increase in CEA only after the rise is confirmed on two separate occasions 2 months apart. This definition is often overlooked in clinical practice, and unfortunately, patients may be thought to have treatment-refractory cancer when they are actually experiencing clinical benefit. Although we now have six drugs approved in the United States for the treatment of advanced colorectal cancer, one should not discard an effective therapy without deriving its maximal benefit for an individual. Many patients derive clinical benefit from a prolongation of time to disease progression without clearly demonstrable radiographic disease shrinkage or progression, a circumstance that can be accompanied by CEA fluctuations in serial testing.
One can also question the appropriateness of continuing therapy for several months to confirm the rising marker, given the toxicity and cost of agents in current use for therapy of advanced disease as well as the availability of alternative assessment measures. In contrast, however, it has been our experience in clinical practice that computed tomography scanning can occasionally be misleading in patients who are responding to multidrug chemotherapy with or without monoclonal antibodies. In some cases with extensive liver disease, we have seen symptomatic responses during dropping CEA counts in the setting of a worsening scan. Persistent treatment with the same agents eventually demonstrated an improvement in the radiographic appearance. In this setting, serial CEA levels may provide evidence to support pressing on with a given therapy.
The bulk of the review focuses on the use of CEA as a means of surveillance after curative-intent surgery for local/regional colorectal cancer. The driving force behind postoperative monitoring is to increase the proportion of patients in whom recurrence is diagnosed early and leads to curative salvage surgery. The authors use their review of the evidence to argue in support of frequent CEA monitoring for this purpose, a recommendation in concert with the ASCO guidelines for postoperative monitoring. Their review reminds us that meta-analysis and randomized trials of intensive postoperative surveillance have shown improved survival as a result of careful monitoring. However, the degree to which CEA contributes is small and somewhat uncertain, as a number of the control arms of these randomized trials included CEA monitoring.
Based on the literature, Drs. Fakih and Padmanabhan report that CEA monitoring alone probably leads to a potentially curative resection in fewer than 5% of all patients undergoing surveillance and, more likely, the true number is in the range of 1% to 3%. If this is an accurate estimate, and assuming 40% are cured by salvage surgery, then CEA testing leads to cure in 0.4% to 1.2% of those who undertake postoperative CEA surveillance.
The cost-effectiveness of the addition of CEA testing to benefit such a small proportion of patients has not been rigorously studied in the era of modern colorectal cancer therapy. However, a model using older therapies estimated its cost-utility to be on the order of $500,000 for each quality-adjusted life-year gained, a figure well beyond the cost of most commonly used oncology interventions. The field is changing quickly enough that the use of CEA testing in this setting needs to be readdressed, especially in light of the recent ASCO guidelines that now support the use of routine abdominal scans as a surveillance tool.
The small benefit must also be placed in perspective: Many patients have CEA results that can be either falsely positive or falsely negative in the surveillance setting. In 1993, Moertel et al reported that the false-negative rate of CEA testing was 41% and the false-positive rate 16%, which translates into a positive-predictive value (the chance that having a positive test means you have colorectal cancer recurrence) of 72% and a negative-predictive value (the chance that having a negative test means you don’t have colorectal cancer recurrence) of 75% in this population. The accompanying editorial by Dr. Fletcher called for an end to the use of CEA in surveillance, given that fairly rigorous trials had repeatedly shown the benefit of CEA to be too small to warrant its further use.
The only significant advance in the field since 1993 has been improved imaging technology that allows for frequent detection of subcentimeter metastases, perhaps even further decreasing the independent utility of CEA. Few of us followed Dr. Fletcher's advice then, but perhaps it is time to reconsider the wisdom of that recommendation.
Despite all of these limitations, CEA testing likely allows quite a small, albeit real, number of patients to initiate a course of events leading to a curative surgical resection of recurrent disease. While there is potential harm caused by false-positive testing and the false sense of security that results from false-negative testing, it is likely outweighed by the desire of both the patient and physician to feel they have done everything possible to have a second chance at cure.
As Drs. Fakir and Padmanabhan note, CEA testing as a means of colorectal cancer surveillance is fairly ineffective and does not make good sense on a population level. However, as clinicians dealing with individuals, we continue to use it in the hope that we might find a cancer early and perhaps give that individual a chance of cure when their disease recurs. A more liberal use of imaging would likely do more to find early disease recurrences than frequent CEA testing, and the new ASCO guidelines published in December 2005 support this concept. As clinicians, we must be cognizant of the significant limitations of CEA testing, and as researchers, we should continue to strive for better markers to alert us that colorectal cancer has recurred. Until then, along with Drs. Fakih and Padmanabhan, we will continue to make judicious use of our imperfect tools.
-Hanna Kelly, MD
-Richard M. Goldberg, MD
Financial Disclosure:The authors have no significant financial interest or other relationship with the manufacturers of any products or providers of any service mentioned in this article.
1. Bast RC Jr, Ravdin P, Hayes DF, et al: 2000 update of recommendations for the use of tumor markers in breast and colorectal cancer: Clinical practice guidelines of the American Society of Clinical Oncology. J Clin Oncol 19:1865-1878, 2001.
2. Desch CE, Benson AB 3rd, Somerfield MR, et al: Colorectal cancer surveillance: 2005 update of an American Society of Clinical Oncology practice guideline. J Clin Oncol 23:8512-8519, 2005.
3. Kievit J, van de Velde CJ: Utility and cost of carcinoembryonic antigen monitoring in colon cancer follow-up evaluation. A Markov analysis. Cancer 65:2580-2587, 1990.
4. Hillner BE, Schrag D, Sargent DJ, et al: Cost-effectiveness projections of oxaliplatin and infusional fluorouracil versus irinotecan and bolus fluorouracil in first-line therapy for metastatic colorectal carcinoma. Cancer 104:1871-1884, 2005.
5. Moertel CG, Fleming TR, Macdonald JS, et al: An evaluation of the carcinoembryonic antigen (CEA) test for monitoring patients with resected colon cancer. JAMA 270:943-947, 1993.
6. Fletcher RH: CEA monitoring after surgery for colorectal cancer. When is the evidence sufficient? JAMA 270:987-988, 1993.