Frequency of Surveillance, CEA Testing an Issue in NCCN Colorectal Cancer Guidelines

April 1, 1996

FORT LAUDERDALE, Fla--The frequency of surveillance after colectomy, particularly CEA testing, was a major topic of discussion after the presentation of the National Comprehensive Cancer Network's (NCCN) preliminary guidelines on colorectal cancer, one of eight such guidelines introduced at the coalition's first annual conference.

FORT LAUDERDALE, Fla--The frequency of surveillance after colectomy,particularly CEA testing, was a major topic of discussion afterthe presentation of the National Comprehensive Cancer Network's(NCCN) preliminary guidelines on colorectal cancer, one of eightsuch guidelines introduced at the coalition's first annual conference.

Paul F. Engstrom, MD, of Fox Chase Cancer Center, who headed thecolorectal cancer guidelines panel, said that monitoring aftercolectomy encompasses more than just CEA testing, and he citedseveral reasons for doing surveillance:

To monitor for complications. "The types oftherapy we're giving are often combined modality with surgery,radiotherapy, and chemotherapy, and a specialist needs to seethese patients often to monitor for complications, at least inthe first 2 years," he said.

To detect curable recurrence. "These are rare,maybe 5% of patients who recur have a curable recurrence,"Dr. Engstrom said, "but we don't want to overlook those patients."

To look for a synchronous or new colon cancer. "Thebest way to prevent these is to remove the premalignant lesions,and that's why colonoscopy is recommended [annually for 2 years,then every 3 years if negative]."

He added two other possible reasons for surveillance: to provideassurance to the patient and to look for other, non-colon cancerrelated malignancies.

In the guidelines, which are still in a preliminary form, CEAtesting is recommended for patients whose CEA level is elevatedat diagnosis or within 1 week of colectomy, at a frequency ofevery 6 months for 2 years, than annually for 5 years.

Criticism of this recommendation came from both sides--those whothought the recommended schedule for CEA testing was not frequentenough and those who questioned the cost effectiveness of performinga test that ultimately benefits only a few patients. One physiciancited Dr. Charles Moertel's retrospective study showing a truebenefit for CEA monitoring in the neighborhood of only 1% or 2%.

Panel member Margaret Tempero, MD, of the University of NebraskaMedical Center, pointed out that the database for this reviewincluded all patients whether or not a preoperative CEA had beendone and whether or not they were CEA positive.

The NCCN recommendation for CEA monitoring, she reminded the audience,is only for patients with tumors that initially express CEA andin whom the CEA returns to normal after resection of the primary.

"I think we're really looking at a focused subset of patientswho are at high risk of recurrence, patients with stage B2 andC tumors that produce CEA," she said.

Dr. Tempero noted that colon cancer is a disease in which metastaticsites can sometimes be resected with potential for cure, "andthat was the rationale for incorporating surveillance into theguidelines." She emphasized that the guidelines are stillevolving and that the frequency of testing is still under debateby the guidelines panel.

A physician from Ohio State said that leading studies have shownthat unless CEA is done every 2 months for the first 2 years,the opportunity to reoperate for cure is lost in 25% of patients,and thus testing only every 6 months may be inappropriate.

Dr. Engstrom agreed that "if you're doing CEAs every 6 months,you're probably going to miss the doubling time that's crucialto a potentially life-saving intervention."

He noted that these are the kinds of issues that the NCCN willstudy through its patterns of care and outcomes research.

Other physicians questioned the utility of treating asymptomaticmetastatic disease, but Dr. Tempero said there is literature tosupport such treatment. A study from the Nordic Cancer TreatmentGroup in which asymptomatic metastatic patients were randomizedto immediate treatment vs delayed treatment showed a slight survivaladvantage for immediate treatment.

Also, she said, new data from a United Kingdom study of colorectalcancer patients with liver metastases, published in The Lancet,showed that patients who were immediately randomized to hepaticarterial infusion therapy had a survival benefit over those assignedto the delayed therapy arm.

Dr. Tempero said that both of these studies provide another argumentfor surveillance, while acknowledging that "we don't havea clear handle on when you should begin treatment in these patients."

Another member of the colon cancer guidelines group, Mark Roh,MD, of M.D. Anderson, emphasized that even repeat hepatic andlung resections may still lead to cures. Based on the literatureand his own practice, he said that a small portion of patientswho have had a repeat hepatic resection for a second metastaticrecurrence have reached the 5-year disease-free survival mark.

"Granted it's a small number," he said, "but enoughto justify doing frequent CT follow-up after resection."

'Was Cost Considered?'

The cost-benefit ratio of surveillance was also an issue for theaudience. When asked if cost was considered in developing theguidelines, Dr. Tempero said it was not. "We were lookingat what is the right thing to do for the patient, so cost didnot enter into the discussion."

An audience member then asked if this meant that the panel wasrecommending "best care no matter the cost," and askedhow this stance could be justified to payers and purchasers.

Dr. Engstrom pointed out that such considerations will be thenext step of the guidelines panels' work. "Do we now haveto take out the 'or's' [alternative treatments] and state whatwe're going to do and stick with it?"

Dr. Tempero noted that all of the guidelines panels are workingwith the outcomes committees to tie in the observation endpointswith the treatment recommendations, in terms of patient functionality,costs of treatment, overall survival, etc.

She also expressed her wish that the cooperative groups and CCOPs(Community Clinical Oncology Programs) take on some of these costof care issues to provide data to help the panels make cost-effectivedecisions.

NCCN Colorectal Cancer Practice Guidelines Panel

Al B. Benson III, MD
Northwestern University

Alfred M. Cohen, MD
Memorial Sloan-Kettering Cancer Center

James H. Doroshow, MD
City of Hope National Meidcal Center

Paul Engstrom, MD
Fox Chase Cancer Center

Krystyna Kiel, MD
Northwestern Memorial Hospital

John E. Niederhuber, MD
Stanford University School of Medicine

Mark Roh, MD
U of Texas M.D. Anderson Cancer Center

Margaret Tempero, MD
University of Nebraska Medical Center