|
Advertisement |
|
|
||
 |
|
|||||
|
|||||
|
 |
||||||
|
||||||
|
||||||
|
© 1999 American Society for Clinical Oncology
Recommended Colorectal Cancer Surveillance Guidelines by the American Society of Clinical OncologyFrom the American Society of Clinical Oncology. (ASCO) Colorectal Cancer Surveillance Panel; C.E.D. and A.B.B. are co-chairs of this panel. Adopted by the ASCO Board of Directors on November 13, 1998. Address reprint requests to Jennifer J. Padberg, American Society of Clinical Oncology, Health Services Research, 225 Reinekers Lane, Suite 650, Alexandria, VA 22314; email padbergj{at}asco.org ABSTRACT OBJECTIVE: To determine the most effective, evidence-based, postoperative surveillance strategy for the detection of recurrent colon and rectal cancer. Tests are to be recommended only if they have an impact on the outcomes listed below. POTENTIAL INTERVENTION: All tests described in the literature for postoperative monitoring were considered. In addition, the data were critically evaluated to determine the optimal frequency of monitoring. OUTCOMES: Outcomes of interest included overall and disease-free survival, quality of life, toxicity reduction, and cost-effectiveness. The American Society of Clinical Oncology (ASCO) Colorectal Cancer Surveillance Expert Panel was guided by the principle of cost minimization, ie, when two strategies were believed to be equally effective, the least expensive test was recommended. EVIDENCE: A complete MEDLINE search was performed of the past 20 years of the medical literature. Keywords included colorectal cancer, follow-up, and carcinoembryonic antigen, as well as the names of the specific tests. The search was broadened by articlesfrom the tumor marker ASCO panel literature search, as well as from bibliographies of selected articles. VALUES: Levels of evidence and guideline grades were rated by a standard process. More weight was given to studies that tested a hypothesis directly relating testing to one of the primary outcomes in a randomized design. BENEFITS/HARMS/COSTS: The possible consequences of false-positive and false-negative tests were considered in evaluating a preference for one of two tests that provide similar information. Cost alone was not a determining factor. RECOMMENDATIONS: The expert panel's recommended postoperative monitoring schema is discussed in this article. VALIDATION: Five outside reviewers, the ASCO Health Services Research Committee, and the ASCO Board of Directors examined this document. SPONSOR: American Society of Clinical Oncology. COLON AND RECTAL cancers were diagnosed in an estimated 131,600 people in the United States in 1998.1 Sixty percent of these patients presented with stage II or III disease, which will recur with metastatic or locally invasive disease in approximately 35% to 40% of them. As a result, hundreds of thousands of people with resected colorectal cancer are candidates for surveillance. The majority of recurrences in patients who have undergone a complete resection of a colorectal cancer will occur within 5 years, and usually within 3 years of surgery. Patients with colon cancer most frequently develop liver metastases as the site of initial recurrence. Rectal cancers, however, may spread through paravertebral venous and lymphatic channels to the lungs, without involving the liver. Because metastatic disease is usually fatal, there has been a tremendous amount of effort focused on finding recurrent colon and rectal cancers before symptoms develop, at a stage when another curative resection is still possible. Carcinoembryonic antigen (CEA) tests, colonoscopies, chest x-rays (CXRs), liver function tests (LFTs), complete blood cell counts (CBCs), fecal occult blood tests (FOBTs), computerized tomography (CT), and ultrasonography were all evaluated in this setting in the hopes of reducing the incidence of incurable metastatic disease. The findings from studies of postoperative monitoring in colorectal cancer have varied widely. As a result of this uncertainty, there is considerable variation in follow-up practice.2 The variation in practice has resulted in wide variation in follow-up costs. For example, the differences between Medicare-allowed charges differed 28-fold (from $561 to $16,492 over a 5-year period).3 As a result of these differences in patterns, costs, and reported outcomes, the American Society of Clinical Oncology (ASCO) convened an expert panel to address the issue of colorectal cancer-related surveillance. PRACTICE GUIDELINES "Practice guidelines are systematically developed strategies to assist practitioner and patient decisions about appropriate health care for specific clinical circumstances."4Attributes of good guidelines include validity, reliability, reproducibility, clinical applicability, clinical flexibility, clarity, multidisciplinary process, review of evidence, and documentation.4 Guidelines may be useful in producing better care and decreasing its cost. Specifically, utilization of clinical guidelines may provide the following: (1) improvements in outcomes, (2) improvements in medical practice, (3) a means for minimizing inappropriate practice variation, (4) decision support tools for practitioners, (5) points of reference for medical orientation and education, (6) criteria for self-evaluation, (7) indicators and criteria for external quality review, (8) assistance with reimbursement and coverage decisions, and (9) criteria for use in credentialing decisions. In formulating recommendations for colorectal cancer surveillance, ASCO considered these tenets of guideline development, emphasizing review of data from controlled clinical trials. However, it is important to realize that guidelines cannot always account for individual variation among patients. They are not intended to supplant physician judgment with respect to particular patients or special clinical situations and cannot be considered inclusive of all proper methods of care or exclusive of other treatments reasonably directed at obtaining the same results. Accordingly, ASCO considers adherence to these guidelines to be voluntary, with the ultimate determination regarding their application to be made by the physician in light of each patient's individual circumstances. In addition, these guidelines describe administration of therapies in clinical practice; they cannot be assumed to apply to interventions performed in the context of clinical trials, given that clinical studies are designed to test innovative and novel therapies in a disease for which better therapy is sorely needed. In that guideline development involves a review and synthesis of the latest literature, a practice guideline also serves to identify important questions for further research and those settings in which investigational therapy should be considered. METHODS A methodology similar to that applied in prior ASCO practice guidelines was used5 and is described in more detail below.
Panel Composition
Process Overview In evaluating the evidence regarding surveillance after primary treatment of colorectal cancer, the panel followed a process for guideline development established by the American College of Chest Physicians.6,7 The process included a systematic weighting of the level of the evidence and a systematic grading of the evidence for making a recommendation (Table 1).6,7
Review of Available Data The panel considered colon and rectal cancer as different tumors where appropriate. However, for most issues, the follow-up is similar. When possible, the tests with the greatest potential to detect recurrence were considered, but alternatives were provided when cost or patient comfort was a major issue. The following outcomes were considered in determining the optimal tests and frequency of postoperative colorectal cancer monitoring: survival, disease-free survival, quality of life, reduction in toxicity, and cost-effectiveness.9 However, these outcomes were difficult to apply in many situations because of the lack of clinical trials addressing specific tests. The panel addressed the success of surgery for isolated metastatic disease. Several published studies show a 20% to 30% long-term survival rate in selected patients undergoing resection for isolated hepatic or pulmonary metastases.10-14 Therefore, the panel also addressed whether any of the tests clearly identified resectable (hepatic, pulmonary, and local) disease before symptoms developed. Until there is better therapy for metastatic disease, extending the lead time (ie, the interval between when the test is positive and when symptoms begin) by frequent testing is unlikely to prolong survival or quality of life. The data are clear, however, that chemotherapy of metastatic colorectal cancer can improve short-term survival and often improve the quality of life.15 Furthermore, there is some evidence that treatment of patients with asymptomatic metastatic colorectal cancer may yield better results than treatment of those patients only once they become symptomatic. The Nordic Gastrointestinal Tumor Adjuvant Therapy Group found that patients treated while asymptomatic had 6 months better average survival, although randomized cohorts were imbalanced in terms of risk.16 Allen-Mersh et al17 found improved survival of 6 months in a group of patients who had received palliative hepatic artery floxuridine compared with those who had not. The National Cancer Institute of Canada is coordinating a confirmatory randomized phase III trial (CAN-NCIC-CO10, NCI-V94-0463) to evaluate immediate versus delayed fluorouracil and leucovorin for asymptomatic advanced colorectal cancer patients. Expected accrual would include 180 patients. The trial is designed to determine whether there are survival and quality-of-life differences for asymptomatic patients who receive chemotherapy before the onset of symptoms from advanced disease compared with those who are treated with chemotherapy only when symptoms develop. The panel focused on the ability to detect those patients with asymptomatic metastatic colorectal cancer and recommended tests that are able to discover a recurrence before symptoms develop and that have been shown consistently in the literature to be the first test to document an isolated metastasis. Because resection of an isolated metastasis seems to be the only way to improve long-term survival for patients with colorectal cancer, most of the commonly available surveillance tests could be justified only in those patients who would be willing and able (ie, lacking significant comorbidity) to undergo a hepatic, pulmonary, or anastomotic resection. The panel limited its attention to sporadic cases of colon and rectal cancer. Patients with hereditable cancers may need more frequent attention; guidelines for screening this population are addressed in the World Health Organization (WHO) publication on the prevention of colorectal cancer.18
Consensus Development Based on Evidence The panel did not attempt to codify established practice. The experts reviewed the available evidence and added their best clinical judgment to make final recommendations, using standardized language to characterize the strength of the evidence. In accordance with the ASCO Health Services Research Policies and Procedures for guidelines, "recommendation" was used when there was level I or II evidence and panel consensus. "Suggestion" was used when there was level III, IV, or V evidence and panel consensus. "No guideline possible" was used when there were no data or the panel could not reach consensus.
Guideline and Conflict of Interest
Revision Dates
Review of Available Data
Yearly colonoscopy detected no asymptomatic local recurrences; only one asymptomatic second primary colon cancer was found. Yearly CT scans detected asymptomatic liver-only metastases much more frequently than the standard screening strategy (10/14 v 0/12, respectively, P = .0002). However, the resection rate of the discovered liver metastases was not significantly different between the two arms (three v four), and only one of these patients has subsequently remained disease-free (intensive arm). The patients allocated to the intensely monitored arm underwent an extra 608 CT scans. The CXR was not very effective at detecting recurrences. Only 5% of patients had asymptomatic lung metastases found by annual CXR, with three patients resectable and only one patient remaining disease-free long term. In the annual screening group, 633 more CXRs were performed. Two other randomized studies compared different levels of follow-up intensity and failed to confirm the benefit of more aggressive testing. Ohlsson et al20 compared an intense follow-up program to a complete absence of routine follow-up tests or visits. The number of recurrences found in each group was the same (32%, or 17/53, and 33%, or 18/54, in the experimental and control groups, respectively). Attempts at curative resections of recurrence were possible in five patients (29%) in the experimental group and in three patients (17%) in the control group. Less than one third of patients with recurring disease, even in the intensely followed group, had a chance of long-term survival with a high-intensity follow-up strategy. Salvage therapy seemed successful in two of five patients in the intensive group (v none of the three patients in the control arm), which is a net 4% of the original population. Intensive surveillance resulted in only a slight improvement in the cancer-specific (78% v 71%) and overall (75% v 67%) 5-year survival rates. Neither of these differences was statistically significant, although the power of the comparisons was limited. In the other prospective randomized trial performed by Makela et al,21 all individuals underwent similar testing with history and physical examination, CEA testing, and CXR. The standard surveillance group underwent rigid sigmoidoscopy (rectal cancer only) and a yearly barium enema. The intensive surveillance group received a flexible sigmoidoscopy every 3 months (rectal cancer only) and a yearly colonoscopy. In addition, intensive surveillance patients underwent liver ultrasonography every 6 months and a yearly abdominal CT scan. Again, recurrence rates were nearly the same in the "intensified follow-up" and control groups (42%, or 22/52, and 39%, or 21/54, respectively). Attempts at curative resection were possible in 22% and 14%, respectively. The results of such surgery were not reported, however. The cancer-specific survival rates at 5 years were not reported, but overall survival rates were not statistically different (59% v 54%). However, as with the study by Ohlsson et al,20 the sample size in the trial reported by Makela et al was relatively small, increasing the chance of a false-negative result. Indeed, there was not enough statistical power in either study to detect a small survival benefit (< 15%) for monitoring. Goldberg et al22 reported on 1,247 patients enrolled in the North American Gastrointestinal Intergroup adjuvant colon cancer clinical trial and followed according to protocol. History and physical examinations, CXRs, LFTs, and CEA testing were performed and hemograms were obtained every 3 months x 4, every 4 months x 3, then every 6 months for a total of 5 years. Proctoscopy and barium enema or colonoscopy were routinely performed at 24 and 40 weeks, then annually. CEA tests were performed on 827 patients. CT scans were not mandated. Of 1,247 patients, 548 (44%) experienced recurrence and 222 (18%) had "curative intent surgery" (CIS) attempted. In 109 patients (9%), CIS could actually be completed. The 5-year disease-free survival rate of the patients who completed CIS was 23%. CIS was possible in 75% of patients detected by surveillance tests, and in only 30% whose recurrence was first signaled by symptoms. For all the surveillance tests performed on 1,247 patients, 28 patients were alive at 5 years after a curative resection of recurrent disease and 12 patients were alive after CIS for a second cancer. The authors wrote that this was a multimillion dollar investment in surveillance and that efforts should be directed toward increasing effectiveness and reducing costs. Full results and motivating factors for CIS are listed in Table 3.
Numerous case series and a quantitative review of nonrandomized trials have suggested that postoperative monitoring affords either a small survival benefit or a more favorable distribution of second curative resections.23-28 However, these studies are plagued by selection bias and a differential intensity of follow-up between the control and intervention groups. None of the studies in the quantitative review was randomized and no specific test was thoroughly examined. Finally, many of the publications addressing the usefulness of a specific test predated the era of adjuvant therapy. Recurrences after adjuvant therapy are less common and perhaps less amenable to treatment (eg, local recurrence of rectal cancer after surgery, chemotherapy, and radiation), which further reduces the "payoff" of surveillance. Understanding the limitations in the scientific data available to address the issue at hand, the following guidelines are presented (Table 4).
COLORECTAL CANCER SURVEILLANCE GUIDELINES
1. Carcinoembryonic Antigen Level of Evidence: II. Grade of Recommendation: C. CEA has been studied extensively, as reported in the literature. The ASCO tumor marker panel guidelines discuss the rationale for postoperative CEA monitoring to detect asymptomatic recurrences,8 and this subject will not be further reviewed here. Of note, approximately 30% of all colorectal cancer recurrences do not produce CEA29; a false-negative CEA test result is more common in poorly differentiated tumors30; and patients with a normal preoperative CEA level may have an elevated CEA level at recurrence (44% in one series).31
2. History and Physical Examination Level of Evidence: V. Grade of Recommendation: Panel Consensus. No formal examination of the contribution of the physician's history and physical examination to the health outcomes of the colorectal cancer patient has been performed. Data from the largest study of surveillance to date showed that nearly 80% of recurrences were found by CEA testing, with only 20% found by a routine history and physical examination done at the same time interval.19 Other studies have confirmed that the history and physical examination added little to the regular laboratory/imaging testing program.21,24,25 Several studies have shown that 45% or more of the recurrences within the first 3 years are prompted by symptoms that occur between physician visits and diagnostic tests.20,26,27 In these studies, regular testing was performed every 3 to 4 months. Although the above-mentioned studies suggest that regular physician visits have only a small direct effect on patient outcomes (ie, recurrences are occasionally detected this way, and some are potentially resectable), these data fail to address other potential benefits of the physician-patient encounter. For example, patients require coordinated care, including scheduling and interpretation of their examination and test results. The physician visit provides an opportunity to provide reassurance and discuss prevention, dietary advice, new therapy, genetic information, and other health concerns. The panel thus suggests clinical follow-up every 3 to 6 months for the first 3 years, then annually. After that, medical follow-up should be driven by symptoms or the need for age-related periodic health maintenance. Current practice is that over 90% of colorectal surgeons follow patients every 3 to 6 months for the first 3 years, and over 80% continue to follow their patients at least yearly.2
3. Liver Function Tests Level of Evidence: IV. Grade of Recommendation: D, Panel Consensus. Most strategies for colorectal cancer follow-up have included LFTs. However, no studies have shown any outcome improvement for the routine monitoring of any of the liver functions. One study clearly shows that results of other blood and radiologic tests become positive well before LFT results do.28 Therefore, the panel agreed that there was no evidence to support the regular monitoring of liver function.
4. Fecal Occult Blood Test Level of Evidence: II. Grade of Recommendation: C. The usefulness of FOBTs for detecting colon cancer recurrence after colon cancer surgery has been compared with the usefulness of structural colorectal evaluation. In the study by Ohlsson et al,20 three recurrences were first found by fecal hemoglobin assay. Two of those recurrences were also discovered during scheduled proctosigmoidoscopy.20 Safi and Beyer29 showed that only 12% of recurrences had mucosal disruption and that the fecal hemoglobin assay was therefore a low-yield test. A trial of serial FOBTs in 1,217 colon cancer survivors demonstrated poor sensitivity and specificity of FOBTs for detecting recurrent cancers or polyps.32 As colonoscopy is recommended by the panel as part of regularly scheduled examinations, according to WHO guidelines,18 the use of FOBTs probably adds little to postoperative monitoring.
5. Computed Tomography Level of Evidence: II. Grade of Recommendation: A. Some published series of postoperative monitoring include abdominal CT or ultrasonography,21,23,26,27 but others do not.20,24,25,28 A large survey of surgical practice shows that approximately 25% of surgeons order a CT scan every 6 to 12 months for 5 years; the majority do not routinely order this test.2 Studies reporting the routine use of liver imaging have been unable to show that this test finds curable metastatic lesions before other tests. Schoemaker et al19 showed that liver metastases were detected in 10 of 14 patients while they were asymptomatic but that only three patients could have curative resection and only one patient was alive at 20 months; these results are consistent with other studies. Makela et al21 showed that CEA testing was the first method to document recurrence in the majority of patients after colorectal surgery. In this study, ultrasonography detected only four of 22 recurrences and CT scanning found two of 22 recurrences before CEA or other testing (which included a CXR, symptoms, endoscopy, and a FOBT). The meta-analysis of colon cancer follow-up showed that routine testing produced a survival benefit; some of the patients included in the analysis had undergone routine ultrasonography, but none had routine CT testing. Only strategies that included a CEA test showed a survival benefit, but the number of liver-imaging tests was small.23 Sugarbaker et al27 examined a series of 66 consecutive patients and showed that liver imaging (including CT scans, liver-spleen scans, and ultrasonography) performed three times a year added little to the benefit of CEA monitoring. The recommended strategy after this study was to perform CEA testing and office visits alone. Deveney and Way26 noted that CT scanning showed the first evidence of recurrence in nine of 23 patients who experienced recurrence, of 65 patients followed. However, seven of the nine had concurrent elevations of their CEA levels. The false-negative rate for CT scanning was 39% and the sensitivity was 61%. Periodic CT scanning has been proposed to detect CEA-negative recurrences. Approximately 30% of all colorectal cancer recurrences do not produce CEA30; a false-negative CEA test result is more common in poorly differentiated tumors.33 However, the usefulness of CT scanning, even in this situation, is unproven. The panel determined that the role of CT scanning, ultrasonography, and magnetic resonance imaging is to diagnose abnormalities prompted by an abnormal CEA test result or clinical symptom. Future studies should directly compare serum CEA and radiologic testing to determine whether there is a role for CT scanning in the asymptomatic, CEA-negative patient.
6. Chest X-Ray Level of Evidence: II. Grade of Recommendation: B. Most follow-up protocols in the literature have included routine CXRs every 6 to 12 months after curative surgery. In a trial that compared yearly CXRs to CXRs only when clinically indicated, 5% of patients had their first recurrence noted on CXRs.19 Only three of 157 patients could be surgically resected, and only one had long-term survival. Most of the other relevant studies are case series (level IV) that did not include a comparison group. Rocklin et al28 showed the CXR was the first study documenting recurrence in five of 17 patients with recurrent disease (level IV). Eight other recurrences detected by CEA testing were located in the liver and pelvis. A randomized study that compared no follow-up with intensive testing, including a CXR, found that the CXR identified two of 17 recurrences in the monitored group and one of 18 in the control group (level II).20 None of the patients underwent a pulmonary resection. Tornqvist et al25 showed that 14 of 69 patients with recurrent disease had pulmonary metastases discovered by CXR only. Seven patients underwent resection and four are free of disease. Another study showed that 15 of 350 recurrences were discovered by yearly CXR.29 Seven patients were resected and only two are free of disease. Arnaud et al34 reported that a surveillance CXR found three of 43 recurrences, but none occurred in a long-term survivor. Of recurrent disease patients followed at the National Institutes of Health, 21% had pulmonary metastases but only 3% of the recurrences were isolated in the lung.27 Despite the low false-positive rate and low cost of CXRs, the advantage in outcome has been small. Although it is the practice of some physicians and cooperative groups to obtain yearly CXRs, the absolute benefit is small (three of 157 screened patients resected, one of 157 alive at 48 months).19
7. Colonoscopy Level of Evidence: I. Grade of Recommendation: B. At the time of the initial colorectal cancer diagnosis, all patients should have visualization of the entire large bowel to determine the presence of metachronous lesions. If colonoscopy of the entire bowel was not (or could not be) performed before resection, a postoperative colonoscopy is warranted. There are no data suggesting that colonoscopy reduces deaths from recurrent cancer. Despite this, direct visualization of the bowel is frequently performed for colon and rectal cancer follow-up. In fact, a survey of colorectal surgeons shows that almost 75% of physicians perform this test every 6 to 12 months for at least 5 years.2 Colonoscopy at 3-year intervals has been shown to have the same outcomes as yearly colonoscopy for patients with polyps.35 Studies on this issue show a wide variation in the results of periodic examinations. A study by Juhl et al36 followed 261 patients after curative surgery with a yearly colonoscopy or barium enema. Forty-four of 56 recurrences occurred within the first 2 years; only one recurrence was detected after 4 years. Most recurrences were in the liver, lung, or abdomen, but nine anastomotic recurrences were found within the first 30 months after surgery. All of these recurrences were in patients who had had a low anterior resection, and none of these patients had totally resectable disease. Four metachronous cancers (all stage A/B and resectable) and 160 polyps were discovered with this screening strategy. In a study by Rocklin et al,28 endoscopy showed the first recurrence in three of 65 patients (17 of the 65 experienced recurrence). In this study, 82% of all recurrences were found within 3 years of the original diagnosis. Using colonoscopy every 6 months for 2 years and then yearly thereafter, Deveney and Way26 followed 65 patients after curative surgery. No recurrences were found solely on the basis of a barium enema, CXR, or colonoscopy. Thirty benign polyps were removed; there were no metachronous cancers. Similarly, another study that used several monitoring tests showed that colonoscopy did not find any of the 35 recurrences (among 107 patients) before some other test documented recurrence, and no metachronous lesions were found.20 In this study, most of the local recurrences were discovered because of symptoms and were located in the pelvis. The study by Tornqvist et al25 showed that only 3% of recurrences were first revealed by barium enema. A larger proportion of recurrences were discovered by proctoscopy (discussed in guideline 8, following). This study was not able to demonstrate any difference in outcomes or resectability in an extremely intense versus less intense screening strategy. A randomized study showed that one of 21 recurrences in a minimal follow-up group, compared with two of 22 recurrences in a more intensely followed group, was discovered by endoscopy.21 At least two of these recurrences were in the pelvis. A quantitative review of nonrandomized trials examined the detection of metachronous cancers in patients with minimal versus intense follow-up. There were 1.7% more cancers discovered in the intense follow-up group. However, the number of curative resections did not differ between the two groups.23 Unfortunately, the method used to detect the metachronous lesions was not identified. Single- or double-contrast barium enemas are alternatives to colonoscopy. The double-contrast method has about a 10% greater positive predictive value over the single-contrast method.37 A prospective study by Norfleet et al38 in 3,006 patients compared barium enema to colonoscopy in patients whose polyps were detected during a fiberoptic sigmoidoscopy. The sensitivities of the single- and double-contrast barium enemas were 13% and 26%, respectively; therefore, colonoscopy was the preferred test in this population. In another study that compared radiologic with direct visualization, the barium enema did not miss polyps, but the false-positive rate was approximately 2%.39 A combination of flexible sigmoidoscopy and double-contrast barium enema has a combined sensitivity of 94% and specificity of 99% for neoplasms.40
8. Flexible Proctosigmoidoscopy (rectal cancer) Level of Evidence: IV. Grade of Recommendation: C, Panel Consensus. Local (pelvic and anastomotic) recurrences are more common in patients with rectal cancer than in those with colon cancer41; however, pre- or postoperative pelvic radiation has reduced the risk of local rectal recurrence to less than 10%. Juhl et al36 showed that all nine anastomotic recurrences were found by endoscopy after a low anterior resection. Most patients reported symptoms before the test was performed, and the tests were not performed as regularly as the protocol suggested. None of the patients could be treated curatively for this problem. Proctosigmoidoscopy was the first test to reveal recurrence in 7% of patients in the study by Tornqvist et al.25 In this series, 16 of 22 patients with anastomotic or pelvic recurrences died after the documentation of recurrence. Six of eight patients who were operated upon for cure underwent a successful resection, but the majority were ultimately not able to be treated for cure. The outcomes of proctosigmoidoscopy on survival, quality of life, and costs have not been directly measured independently of other follow-up testing. However, the panel was convinced that successful surgery for an anastomotic recurrence for patients who had not received pelvic radiation would enhance survival and quality of life. Therefore, the panel suggests periodic direct examination of the rectum only for patients who have not received pelvic radiation. The panel underscores, however, that combined chemotherapy and pelvic radiation represent the standard treatment for patients with stage II and stage III rectal cancer. The WHO surveillance guidelines18 devised for patients with adenomatous polyps should be followed for patients who have received pelvic radiation.
9. Pelvic Imaging Level of Evidence: IV. Grade of Recommendation: D. Ohlsson et al20 performed pelvic CT scanning on every patient at 3, 6, 12, 18, and 24 months after an anteroposterior resection. The pelvis was the initial site of relapse in four of six patients. However, all four patients were detected because of the appearance symptoms; the pelvic CT scan simply corroborated the recurrence and did not signal recurrent cancer. Makela et al21 showed that CT scanning revealed the first abnormality in only two of the 19 patients with recurrence in the pelvis. In neither of these cases did the CT scan prompt a curative resection. Pre- or postoperative pelvic radiation has reduced the risk of local recurrence to less than 10%. The panel recognizes the impact pelvic recurrences can have on quality of life. For instance, if further treatment were available in a patient who had not had surgery, radiation, and chemotherapy, detecting a recurrence before symptoms appear might result in a useful intervention. Therefore, the panel suggests CT imaging for patients who may benefit from a treatment that could provide a quantitative or qualitative benefit. The optimal duration of CT imaging in this situation has not been proven; the expert panel believes that for those patients who have had local excision or resection without pelvic radiation, more than 80% of local recurrences would be detected within the first 36 months. If the patient has already received surgery, radiation, and chemotherapy, however, detecting a local recurrence before the appearance of symptoms is not likely to change outcome or treatment planning. Furthermore, it is often difficult to detect local recurrence by pelvic CT in patients who have had surgery and pelvic radiation.
10. Complete Blood Cell Count Level of Evidence: V. Grade of Recommendation: Panel Consensus. There was no formal literature comparing postoperative monitoring of any portion of the CBC to other methods of colorectal follow-up. Because other tests have greater sensitivity and specificity than the CBC for this function, and because there is no biologic relationship between hemoglobin levels and colorectal cancer recurrence, the panel agreed that there was no clinical indication to recommend this test in the absence of bleeding or infection. FUTURE DIRECTIONS A guideline should serve as an indicator of good practice as determined by the best evidence. It should not be used to codify current clinical practice or create new practice patterns without objective evidence. The guideline process should communicate that there is either sufficient evidence to either support or recommend against a particular posttreatment follow-up regimen, or communicate that no guideline is possible based on existing data. In the absence of published evidence, it is important to propose a research plan that would help determine what test(s) or procedure(s) can be objectively justified. The quality of the cancer-related follow-up literature is poor. Good-quality clinical trials are desperately needed to sort out which tests make a difference in important outcomes. Most physicians and patients believe that follow-up will provide benefit and better outcomes. However, this issue needs rigorous study to separate wishful thinking from clinical fact. The panel believes that the most important area to study at this point is the usefulness of radiologic (CT) follow-up of colon and rectal cancer. A study that randomized patients with routine CEA testing to CT versus no CT to follow resectability and survival could clarify the impact of CT testing on patient outcomes and cost. The treatment of colorectal cancer is evolving, particularly with the introduction of new agents. Most of the studies that have failed to show a survival benefit with intensive surveillance were performed before the introduction of optimal biochemical modulation of fluorouracil therapy and before the introduction of irinotecan. More recent data confirm that chemotherapy in metastatic colorectal cancer can improve short-term survival and quality of life.15,42,43 Treatment of patients with asymptomatic colorectal cancer may yield better results than treatment of those patients only after they become symptomatic. Ongoing and future clinical trials will help determine which specific tests are optimal for the detection of early asymptomatic metastases.
ACKNOWLEDGMENTS The expert panel expresses its gratitude to Drs Daniel G. Haller, Robert J. Mayer, and Margaret A. Tempero for their thoughtful reviews of earlier versions of the guidelines. REFERENCES 1. American Cancer Society: Cancer Facts & Figures 1996. Atlanta, GA, American Cancer Society, 1996 2. Virgo KS, Wade TP, Longo WE, et al: Surveillance after curative colon cancer resection: Practice patterns of surgical subspecialists. Ann Surg 2:472-482, 1995 3. Virgo KS, Vernava AM, Long WE, et al: Cost of patient follow-up after potentially curative colorectal cancer treatment. JAMA 273:1837-1841, 1995[Abstract] 4. Anonymous: The Canadian Task Force on the Periodic Health Examination. Can Med Assoc J 121:1193-1254, 1979[Medline]
5.
American Society of Clinical Oncology: American Society of Clinical Oncology recommendations for the use of hematopoietic colony-stimulating factors: Evidence-based, clinical practice guidelines. J Clin Oncol 12:2471-2508, 1994 6. Cook DL, Guyatt GH, Laupacis A, et al: Rules of evidence and clinical recommendations on the use of antithrombotic agents. Chest 102:305S-311S, 1992 (suppl 4) 7. Sackett DL: Rules of evidence and clinical recommendations on the use of antithrombotic agents. Chest 95:2S-4S, 1989 (suppl 2)
8.
American Society of Clinical Oncology: Clinical practice guidelines for the use of tumor markers in breast and colorectal cancer. J Clin Oncol 14:2843-2877, 1996
9.
American Society of Clinical Oncology Outcomes Working Group: Outcomes of cancer treatment for technology assessment and cancer treatment guidelines. J Clin Oncol 14:671-679, 1996 10. Hughes KS, Simon R, Songhorabodi S: Resection of the liver for colorectal carcinoma metastases: A multi-institutional study of patterns of recurrence. Surgery 100:278-284, 1986[Medline] 11. Scheele J, Stangl R, Altendorf-Hofmann A, et al: Indicators of prognosis after hepatic resection for colorectal secondaries. Surgery 110:13-29, 1991[Medline] 12. Steele G, Bleday R, Mayer RJ, et al: A prospective evaluation of hepatic resection for colorectal carcinoma metastases to the liver. J Clin Oncol 9:1105-1112, 1991[Abstract] 13. Fuhrman GM, Curley SA, Hohn DE, et al: Improved survival after resection of colorectal liver metastases. Ann Surg Oncol 6:537-541, 1995 14. McCormack PM, Burt ME, Bains MS, et al: Lung resection for colorectal metastases 10-year results. Arch Surg 127:1403-1406, 1992[Abstract] 15. Cunningham D, Pyrhonen S, James RD, et al: Randomised trial of irinotecan plus supportive care versus supportive care alone after fluorouracil failure for patients with metastatic colorectal cancer. Lancet 352:1413-1418, 1998[Medline] 16. Nordic Gastrointestinal Tumor Adjuvant Therapy Group: Expectancy or primary chemotherapy in patients with advanced asymptomatic colorectal cancer: A randomized trial. J Clin Oncol 10:904-911, 1992[Abstract] 17. Allen-Mersh TG, Earlam S, Fordy C, et al: Quality of life and survival with continuous hepatic-artery floxuridine infusion for colorectal liver metastases. Lancet 344:1255-1260, 1994[Medline] 18. Winawer SJ, St John DJ, Bond JH, et al: Prevention of colorectal cancer: Guidelines based on new data: WHO Collaborative Center for the Prevention of Colorectal Cancer Bull World Health Organ 73:7-10, 1995[Medline] 19. Schoemaker D, Black R, Giles L, et al: Yearly colonoscopy, liver CT and chest radiograph do not influence 5-year survival of colorectal cancer patients. Gastroenterology 114:7-14, 1998[Medline] 20. Ohlsson B, Breland U, Ekberg H, et al: Follow-up after curative surgery for colorectal carcinoma. Dis Colon Rectum 38:619-626, 1995[Medline] 21. Makela JT, Laitinen SO, Kairaluoma MI: Five-year follow-up after radical surgery for colorectal cancer: Results of a prospective randomized trial. Arch Surg 130:1062-1067, 1995[Abstract]
22.
Goldberg RM, Fleming TR, Tangen CM, et al: Surgery for recurrent colon cancer: Strategies for identifying resectable recurrence and success rates after resection—Eastern Cooperative Oncology Group, the North Central Cancer Treatment Group, and the Southwest Oncology Group. Ann Intern Med 129:27-35, 1998 23. Bruinvels DJ, Stiggelbout AM, Kievit J, et al: Follow-up of patients with colorectal cancer: A meta-analysis. Ann Surg 219:174-182, 1994[Medline] 24. Ovaska J, Jarvinen H, Kujara H, et al: Follow-up of patients operated on for colorectal carcinoma. Am J Surg 159:593-596, 1990[Medline] 25. Tornqvist A, Ekelund G, Leandoer L: The value of intensive follow-up after curative resection for colorectal carcinoma. Br J Surg 69:725-728, 1982[Medline] 26. Deveney KE, Way LW: Follow-up of patients with colorectal cancer. Am J Surg 148:717-722, 1984[Medline] 27. Sugarbaker PH, Gianola FJ, Dwyer A, et al: A simplified plan for follow-up of patients with colon and rectal cancer supported by prospective studies of laboratory and radiologic test results. Surgery 102:79-87, 1987[Medline] 28. Rocklin MS, Slomski CA, Watne AL: Postoperative surveillance of patients with carcinoma of the colon and rectum. Am Surg 1:22-27, 1990 29. Safi F, Beyer HG: The value of follow-up after curative surgery of colorectal carcinoma. Cancer Detect Prev 17:417-424, 1993[Medline] 30. Fletcher RH: Carcinoembryonic antigen. Ann Intern Med 104:66-73, 1986 31. Zeng Z, Cohen AM, Urmacher C: Usefulness of carcinoembryonic antigen monitoring despite normal preoperative values in node-positive colon cancer patients. Dis Colon Rectum 36:1063-1068, 1993[Medline] 32. Ahlquist DA, Wieand HS, Moertel CG, et al: Accuracy of fecal occult blood screening for colorectal neoplasia. JAMA 269:1262-1267, 1993[Abstract] 33. Goslin R, Steele G, Macintyre J, et al: The use of preoperative plasma CEA levels for the stratification of patients after curative resection of colorectal cancers. Ann Surg 192:747-751, 1990 34. Arnaud JP, Bergamaschi R, Casa C, et al: The rationale for CEA dosage in the follow-up of patients operated for colorectal cancer: A prospective study on 800 cases. Spec Int Col Mtg Surg Oncol 25, 1992
35.
Winawer SJ, Zauber AG, O'Brien MJ, et al: Randomized comparison of surveillance intervals after colonoscopic removal of newly diagnosed adenomatous polyps: The National Polyp Study Workgroup. N Engl J Med 328:901-906, 1993 36. Juhl G, Larson GM, Mullins R, et al: Six-year results of annual colonoscopy after resection of colorectal cancer. World J Surg 14:255-261, 1990[Medline]
37.
Ott DJ, Scharling ES, Chen YM, et al: Positive predictive value and post test probability of diagnosis of colonic polyp on single and double contrast barium enema. AJR Am J Roentgenol 153:745-739, 1989 38. Norfleet RG, Ryan ME, Wyamn JB, et al: Barium enema versus colonoscopy for patients with polyps found during flexible sigmoidoscopy. Gastrointest Endosc 37:531-534, 1991[Medline] 39. Stein W, Farina A, Gaffney K, et al: Characteristics of colon cancer at the time of presentation. Fam Pract Res J 13:355-363, 1993[Medline] 40. Jensen AB, Madsen B, Anderson P, et al: Information for cancer patients entering a clinical trial: An evaluation of an information strategy. Eur J Cancer 29A:2235-2238, 1993 41. Minsky BD, Mies C, Recht A: Resectable adenocarcinoma of the rectosigmoid and rectum: Patterns of failure and survival. Cancer 61:1408-1416, 1988[Medline] 42. Rougier P, Van Cutsem E, Bajetta E, et al: Randomised trial of irinotecan versus fluorouracil by continuous infusion after fluorouracil failure in patients with metastatic colorectal cancer. Lancet 352:1407-1412, 1998[Medline] 43. Van Cutsem E, Bajetta E, Nierderle N, et al: A phase III multicenter randomized trial comparing CPT-11 to infusional 5FU in patients (Pts) with advanced colorectal cancer (ACRC) after 5FU failure. Proc Am Soc Clin Oncol 17:256a, 1998 (abstr 984) Submitted February 2, 1999; accepted February 2, 1999. This article has been cited by other articles:
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
2 years after diagnosis. An elevated CEA level, if confirmed by retesting, warrants further evaluation for metastatic disease but does not justify the institution of systemic therapy for presumed metastatic disease. 
