Colorectal Cancer Survivorship: Follow-Up Questions

Most survivorship pages imply that a follow-up schedule is settled science. NCI's health-professional summary says the opposite. Its opening line on the subject is blunt. Limited data, and no high-level evidence, are available to guide patients and physicians on surveillance after surgery and adjuvant therapy.

The American Society of Clinical Oncology and the National Comprehensive Cancer Network each recommend specific surveillance strategies. But the PDQ adds points that are rarely quoted. Periodic evaluations may lead to finding recurrent disease earlier. That monitoring has limited effect on overall mortality. The reason is that few localized, potentially curable metastases are found in people with recurrent colon cancer. And to date, no large-scale randomized trial has shown an overall survival benefit for standard monitoring programs after surgery.

The summary goes further. The best regimen and frequency of follow-up examinations are not well defined. The effect on survival is unclear, and the quality of the data is poor.

None of that means skipping follow-up. It means the schedule is a judgment call built on guideline consensus. It is not a proven survival treatment. That is worth knowing before treating one missed appointment as a disaster. It is also worth knowing before assuming more scans must mean better outcomes.

Carcinoembryonic antigen (CEA) is a serum glycoprotein used often in colon cancer management. The PDQ cites a review of the marker with three specific conclusions:

• A CEA level is not a valuable screening test for colorectal cancer, because of the large number of false-positive and false-negative reports.
• Postoperative CEA testing should be restricted to people who would be candidates for resection of liver or lung metastases.
• Routine use of CEA levels alone for monitoring response to treatment is not recommended.

That second point is the one that changes conversations. The logic is surgical. The reason to catch a rising CEA is that something can be done about it. What can be done is removing a liver or lung metastasis. For someone who would not be a candidate for that surgery, the role of the test is different. It is fair to ask directly why it is being drawn.

CEA does one thing clearly. An elevated CEA before treatment carries negative prognostic significance.

Prognosis in colon cancer rests on three features, and these form the basis of every staging system for the disease:

• How far the tumor penetrated through the bowel wall.
• Whether lymph nodes were involved.
• Whether distant metastases are present.

Two events at presentation are separately listed as indicators of poor prognosis: bowel obstruction and bowel perforation.

One molecular finding cuts the other way. Microsatellite instability is also linked to hereditary nonpolyposis colorectal cancer. It has been associated with better survival, independent of tumor stage. That came from a population-based series of 607 patients younger than 50 with colorectal cancer. Many other markers have been looked at after the fact. Those include allelic loss of chromosome 18q and thymidylate synthase expression. Most have not been validated in advance.

Age is not itself a treatment factor. The PDQ states that treatment decisions depend on physician and patient preferences and on stage, rather than on the patient's age.

A meta-analysis of seven prospective cohort studies looked at physical activity before and after a colorectal cancer diagnosis. Two comparisons stand out.

People who did any physical activity after diagnosis, compared with none, had a relative risk of 0.74 for colorectal cancer-specific mortality. The 95% confidence interval ran 0.58 to 0.95, with P = .02. People doing a high amount of activity, versus a low amount, had a relative risk of 0.65. That interval ran 0.47 to 0.92, with P = .01.

Note what the comparison is. The first figure is any activity against none. That is a lower bar than most exercise advice implies.

Two prospective observational studies drew on patients enrolled in CALGB-89803, an adjuvant chemotherapy trial for stage III colon cancer.

Patients in the lowest fifth of a Western dietary pattern were compared with the highest fifth. Their adjusted hazard ratio for disease-free survival was 3.25 (95% CI, 2.04–5.19; P < .001). For overall survival it was 2.32 (95% CI, 1.36–3.96; P < .001). Separately, stage III patients in the highest fifth of dietary glycemic load had an adjusted hazard ratio for overall survival of 1.76 (95% CI, 1.22–2.54; P < .001), against the lowest fifth.

The Cancer Prevention Study II Nutrition Cohort followed 2,315 people diagnosed with colorectal cancer. Red and processed meat intake before diagnosis was linked to a higher risk of death. The relative risk was 1.29 (95% CI, 1.05–1.59; P = .03). Red meat eaten after diagnosis was not linked to overall mortality.

That before-and-after split is unusual, and worth carrying into a conversation. It complicates the instinct that changing diet after treatment must be the lever.

The PDQ attaches a warning to all of it. Cohort studies have suggested diet or exercise may improve outcomes. No prospective randomized trials have confirmed those findings. The cohort studies had many openings for unintended bias, and caution is needed when using their data.

A prospective cohort study examined aspirin use after a colorectal cancer diagnosis. Regular users had a hazard ratio for colon cancer-specific mortality of 0.71 (95% CI, 0.53–0.95) and for overall mortality of 0.79 (95% CI, 0.65–0.97).

Then it gets more specific. One study looked at 964 patients with rectal or colon cancer. They came from the Nurses Health Study and the Health Professionals Follow-up Study. Among those carrying PI3K variants, regular aspirin use was linked to a hazard ratio for death from any cause of 0.54 (95% CI, 0.31–0.94; P = .01).

All of that is graded as cohort-level evidence, not trial evidence. Aspirin also carries bleeding risk. So this is a question for the treating team, not a conclusion to act on alone. What makes it worth raising is the PI3K angle. That is a concrete, testable molecular question rather than general lifestyle advice.

• Which guideline is this surveillance schedule following — ASCO, NCCN, or something else — and what does it specify for this stage?
• Would I be a candidate for surgery to remove a liver or lung metastasis if one were found? If not, what is the CEA for?
• What was my pretreatment CEA, and what is it now?
• What did the pathology report say about depth of invasion through the bowel wall, and how many nodes were examined and involved?
• Was there obstruction or perforation at presentation?
• Was microsatellite instability or mismatch repair testing done, and what did it show?
• Was PI3K status tested, and is aspirin worth discussing given my bleeding risk?
• When is the next colonoscopy, and who is responsible for scheduling it?

Related pages: colorectal cancer covers the disease itself. What a survivorship care plan is covers how these schedules get written down and shared with primary care. And fear of recurrence covers the part surveillance does not treat.

• NCI PDQ — Colon Cancer Treatment, health professional version