The Preoperative Assessment And

Surg Clin N Am 82 (2002) 1091–1108

The preoperative assessment and postoperative surveillance of patients with colon and rectal cancer Walter E. Longo, MD*, Frank E. Johnson, MD Department of Surgery, St. Louis University School of Medicine, St. Louis, MO 63110, USA

Colorectal cancer is a relatively common malignancy, particularly in developed countries in the Western world. Once diagnosed, it is essential to implement adequate preoperative staging, to devise a treatment plan that minimizes the risk to the patient, and, once treatment is complete, to monitor the patient for the development of recurrent cancer or the development of a new primary. The safety of colectomy and proctectomy has been improved by advances in surgical technique, anesthesia, intensive care therapy, and other supportive measures. Furthermore, resectability rates have steadily increased and the operative mortality has decreased. Historically, the estimates of the risks of adverse outcomes have been based on univariate and bivariate analyses of possible risk factors, such as type of procedure or principal diagnosis [1]. Most of these have been retrospective in nature, although the National Veterans Affairs Surgical Quality Improvement Program (NSQIP) has risk stratified outcomes prospectively [2]. Efforts to minimize the risk of surgery following resection for colorectal cancer will continue to evolve. The aim of follow-up is early detection of recurrence or the discovery of a new primary. The key elements are patient-doctor contact, serum tumor markers, flexible endoscopy, and diagnostic imaging [3]. Today, physicians have at their disposal a variety of biochemical and imaging modalities for the effort to detect these recurrences while they are potentially curable. It remains a real dilemma, both practically and economically, whether any specific test or combination of tests offers the best results. The aim of this article

* Corresponding author. St. Louis University Hospital, 3635 Vista Avenue, St. Louis, MO 63110-0250. E-mail address: [email protected] (W.E. Longo). 0039-6109/02/$ - see front matter
2002, Elsevier Science (USA). All rights reserved. PII: S 0 0 3 9 - 6 1 0 9 ( 0 2 ) 0 0 0 5 0 - 6

1092

W.E. Longo, F.E. Johnson / Surg Clin N Am 82 (2002) 1091–1108

is to reaquaint the physician with the preoperative evaluation and postoperative surveillance of the patient diagnosed with colorectal cancer. Preoperative evaluation Preoperative assessment Once the diagnosis of colorectal cancer has been made, it is necessary to make a judgment on the most appropriate form of treatment for the patient. For a small group of patients, nonoperative therapy will be in their best interest, due to limited survival from pre-existing comorbidities or far advanced neoplastic disease. Most cancers, however, are treated by surgical excision in an effort to achieve cure. The resection may be performed via a transabdominal route, or may be a local excision done transanally for early rectal tumors. If the tumor is considered to be incurable, palliation may require either resection or bypass. Intraluminal stents are defining their role in palliative situations. Only with far advanced metastatic disease is laparotomy or some form of surgery not performed. Before focusing on the malignancy, it remains essential to assess age and pre-existing illnesses. Elderly patients commonly have associated cardiovascular and respiratory problems, or some other comorbid condition that may influence a decision on their suitability for surgery. A well-conducted clinical examination is the most important initial step in evaluating patients for not only occult comorbidities, but also the extent of disease. The abdominal examination may reveal massive abdominal distention due to an obstructing tumor, ascites, or a mass. A liver with associated tumor metastases may be palpated. A digital rectal examination enables assessment of the location of a rectal cancer and its distance from the anal margin, as well as the extent of involvement of the circumference of the rectal wall. Mobility or fixation of the tumor is assessed to determine suitability of local treatment. Tumors in the lower third of the rectum may have caused secondary involvement of the inguinal lymph nodes [4]. If colonoscopy has not been used to make the diagnosis, its value as part of the preoperative evaluation makes it imperative. It is used to detect synchronous lesions, either second primaries or synchronous polyps. It may be performed the day before planned surgery to obviate the need for a second bowel preparation. Pathologic confirmation should be identified so as not to be misled by the rectal mass, which is prostatic in origin, or the cecal lesion, which may be nothing more than a lipoma. A chest radiograph is required to determine if there is coexisting chest morbidity or distant metastases. Urologic evaluation is often a priority for surgeons dealing with rectal tumors; however computed tomography (CT) has virtually replaced the intravenous pyelogram. A complete blood count, liver chemistries, an assessment of renal function, a coagulation profile, and an electrocardiogram are essential laboratory tests. Carcinoembryonic Antigen (CEA) is of limited value in screening for

W.E. Longo, F.E. Johnson / Surg Clin N Am 82 (2002) 1091–1108

1093

colorectal carcinoma. In patients with newly diagnosed disease, preoperative CEA level has been applied usefully in assessing prognosis, and early recurrence and depth of invasion. Nevertheless, the primary application of CEA still remains in the postoperative setting, where increasing values suggest recurrence and persistently normal values suggest the absence of recurrence. Preoperative imaging Computed tomograhy is used to determine the presence or absence of clinically undetectable distant metastases. It rarely causes the surgical plan to be altered, however. One study demonstrated that among 158 consecutive patients with primary colorectal carcinoma, 88 patients had 120 findings present on CT that were otherwise unknown [5]. More important, 35% were significant enough to alter therapy. CT has also been used to assess the stage of the primary tumor, although in the absence of advanced disease, its accuracy remains in question. Regardless of the information reported in the literature either supporting or refuting the routine use of CT scanning in the preoperative evaluation of the newly diagnosed patient with colorectal cancer, most surgeons will obtain this study before operative therapy. Endorectal ultrasound (ERUS) is currently a well-established method and appears to be the gold standard for the preoperative assessment of most rectal cancers. It is fast, minimally invasive, portable, and rapidly interpretable. Currently, the accuracy of ERUS in evaluating perirectal neoplastic infiltration varies between 81% and 96%, whereas the accuracy in demonstrating perirectal lymph node involvement ranges between 60% and 83% [6]. These techniques are obviously operator dependent and results should be evaluated in relation to the experience of the sonographer. The position of the tumor may provide pitfalls in endoluminal scanning. High rectal lesions represent a particular problem with this technique because of poor acoustic coupling and difficulty negotiating the probe alongside the tumor leading edge. To increase the value of ERUS in the staging of stenotic tumors, water enema transvaginal ultrasound may be employed with high sensitivity and specificity. The advent of positron emission tomography (PET) has made it possible to demonstrate chemical and metabolic changes associated with various disease processes. PET will define fundamental differences in the metabolism of F-18-labeled-2 deoxyglucose (FDG) by malignant cells when compared with normal cells. Although PET has acceptable sensitivity in the preoperative diagnosis of occult hepatic metastatic disease, it may find a specific use in patients in whom CT scanning or magnetic resonance imaging (MRI) are suggestive of recurrence. Because PET provides a functional image rather than a precise anatomical definition of the tumor, it lacks specificity and anatomical orientation. Whole-body PET was performed preoperatively in 24 patients with primary colorectal cancer, and the results were compared with histopathology [7]. The PET positive rate was 95.8% for the primary

1094

W.E. Longo, F.E. Johnson / Surg Clin N Am 82 (2002) 1091–1108

tumor. Among 9 patients with histologically confirmed lymph node metastases, 2 (22%) were positive. There were two false positives among the 15 with node negative disease histologically. These results suggest that preoperative PET is useful for the diagnosis of primary colorectal cancer, but is limited in value for detecting metastases to the regional lymph nodes. To date, there is little evidence for the routine use of a role for PET in the preoperative staging of primary colorectal cancer. Radioimmunoscintigraphy is a functional examination that allows for the in vivo imaging of abnormal pathologic processes such as infections, infarcts, and malignant tumors. The recent production of murine and humanized monoclonal antibodies that specifically target well-characterized epitpes located on tumor-associated carcinoembryonic antigen has permitted their use as diagnostic agents in radioimmunoscintigraphy and radioimmunoguided surgery (RIGS). The main use of RIGS appears to be in the preoperative detection of unsuspected advanced locoregional disease and as a complement to CT or MRI imaging in equivocal cases. Below is a suggested preoperative assessment: • • • • • • • •

Assessment of operative risk Complete blood count Coagulation studies Hepatic and renal function Serum tumor markers Colonoscopy Abodominopelvic CT scanning Endoluyminal ultrasound (rectal cancer)

Assessment of operative risk The surgical mortality rates related to colorectal surgery clearly depend on patient selection and underlying pathology. The main causes of death in most prior series have been myocardial infarction, bronchopneumonia, pulmonary embolism, ananstomotic leakage, and cerebrovascular accidents [8]. Outcomes after resection for colon and rectal cancer have been extensively reported, but are difficult to interpret because of study design. Operative mortality rates appear to have plateaued at 1% to 6%, but these results have been achieved in specialized centers where emergency cases have often been excluded and indigent, poorly educated, and noncompliant patients are typically underrepresented. There are a number of factors that have impacted on outcome, such as whether the operation is elective or emergent, the patient’s age, and pre-existing comorbid conditions. The experience of the operating surgeon has also been correlated with patient mortality rates. Radical resection for carcinoma of the rectum is associated with many of the same complications as in other types of intra-abdominal surgery, plus specific genitourinary and stomal complications. Prior studies have indicated that the causes of postsurgical death after proctectomy are varied,

W.E. Longo, F.E. Johnson / Surg Clin N Am 82 (2002) 1091–1108

1095

with cardiopulmonary problems and sepsis dominating the list. The overall mortality rates reported from the St. Mary’s Large Bowel Cancer Project found 6.3% mortality among 4225 patients undergoing either low anterior resection or abdominoperineal resection [9]. Reported morbidity rates associated with Abdominoperineal Resection (APR) range from 34% to 74%, with the most common complications involving the genitourinary tract, poor healing of the perineal wound, and stomal complications. In patients undergoing sphincter-saving procedures, reported anastomotic leak rates range from 0% to 16%. Conclusions regarding the risk factors that accounted for the observed complications are based on weak evidence. The National VA Surgical Quality Improvement Program consists of a prospective registry of more than 1,000,000 patients undergoing major surgical operations [3]. One of the aims of the NSQIP is the development and validation of models that can predict morbidity and mortality for individual patients. Because there are frequent, unpredictable incremental advances in surgical technique and patient care, it is important that risk modeling be done based on large recent patient populations. The NSQIP database is a current, constantly enlarging database that will allow future analyses to focus sharply on details of risk modeling with increased statistical power. Using NSQIP data, previous studies have demonstrated that morbidity and mortality rates for patients undergoing colectomy or proctectomy for cancer are well within the range seen in non-Veterans Affairs institutions [10,11]. Among 5853 patients who underwent colectomy for cancer, 28% developed one or more complications where prolonged ileus, pneumonia, failure to wean from the ventilator, and urinary tract infections were the most frequent complications. The 30-day mortality rate was 5.7%. For most complications, 30-day, in-hospital mortality rates were significantly higher for patients with a complication than for those without. Among 491 patients undergoing proctectomy for cancer, the 30-day mortality rate was 3.2%. Thirty percent of patients had one or more complications after proctectomy. Prolonged ileus, urinary tract infections, pneumonia, and deep wound infection were the most frequently reported complications. Similar to the colectomy data, 30-day mortality rates were significantly higher for proctectomy patients with complications than for those without. The preoperative risk factors that predicted a high risk of 30-day mortality for the colectomy and proctectomy patients are listed below: • Colectomy • Ascites • hypernatremia • Do Not Resuscitate (DNR) status • Low serum albumin • Proctectomy • Elevated Blood Urea Nitrogen (BUN) • Impaired sensorium

1096

W.E. Longo, F.E. Johnson / Surg Clin N Am 82 (2002) 1091–1108

• Low serum albumin • Prolonged Partial Thromboplastin Time (PTT) Postoperative surveillance The main aims of follow-up programs include: (1) to measure the efficacy of the original therapy; (2) to detect metachronous or new malignancies; and (3) to detect potentially curable recurrence [12]. Some argue that the only irrefutable benefit of postoperative follow-up at this time is in the identification of metachronous disease such as adenomatous polyps, new colorectal malignancies, and new primaries in other organ systems. A number of reports have now documented clear benefit in terms of survival duration and cure rates in highly selected patients following resection of recurrent colorectal disease. These reports provide a rationale for postoperative follow-up, but the assumption that more intensive postoperative surveillance is more beneficial than less intensive surveillance is controversial. No study has definitively shown that aggressive postoperative follow-up leads to prolonged disease-free survival or cure. Indeed, several studies have brought into question the usefulness of any follow-up regimen [13–15]. Regardless, modern imaging techniques have led to improved detection of recurrent disease, especially in asymptomatic patients. The recurrence of colorectal cancer is for the most part a time-limited phenomenon. Sixty to 80% of recurrences become apparent within the first two years after initial resection, and 90% within the first four years [14–16]. At least 20% of people with a history of colorectal carcinoma will develop hepatic metastases. Of these, around 15% to 20% will have disease confined to a localized volume within the liver amenable to cure [17]. Following successful resection of an isolated hepatic metastasis, five-year survival rates in large multi-institutional studies range from 20% to 30%, with operative mortality rates of less than 5%. The mean survival duration with untreated hepatic metastases from colorectal carcinoma is approximately six months [18]. Approximately 85% of patients in the United States with colorectal cancer now receive surgery with curative intent. Nevertheless, among the 150,000 patients who receive a diagnosis of large bowel cancer annually in the United States, approximately 60,000 will die of their disease, including many initially treated with curative aim. Potentially curative salvage therapy has proven successful in the minority of patients with recurrent colorectal cancer, particularly those with liver metastases and lung metastases. Successful curative therapy for locally recurrent rectal cancer has also been reported. The potential for cure of a second, newly discovered colorectal cancer is always possible and suggests continued surveillance following cure of the first primary. As surgeons often follow their patients after curative resection of colorectal cancer, an aggressive posture is taken regarding surveillance, realizing that earlier diagnosis of a recurrence or a second primary may allow for more potential therapeutic options. Along the same lines, cost

W.E. Longo, F.E. Johnson / Surg Clin N Am 82 (2002) 1091–1108

1097

effectiveness in a health-care market plagued with financial woes will have an impact on the gatekeepers who authorize the myriad of tests required to look for recurrence or a new primary. Methods of recurrence detection History and physical examination A good history and meticulous physical examination for the patient who has been treated for a previous malignancy cannot be underestimated. Information and education regarding the patient’s tumor biology should be transmitted to the primary care physician, who may see the patient for other co-existing illnesses. Beart et al reported on 168 colorectal cancer patients who were followed prospectively after primary resection [19]. Each patient was seen in follow-up a minimum of every four months and various diagnostic studies were used to complement the history and physical examination. Of the 168 patients followed, 48 developed recurrence. Of these, 41 had symptoms prior to confirmation of recurrence by physical examination finding, serologic testing, or radiologic examination. These symptoms included coughing, abdominal or pelvic pain, change in bowel habits, rectal bleeding, and malaise. Physical examination was less sensitive than patient history, suggesting that the report of symptoms may lead to earlier diagnosis of recurrence than frequent physical examinations. Those with positive findings had either symptoms or elevated CEA levels. It is generally believed, but not proven, that recurrent disease that is symptomatic or can be detected on physical examination is more likely to be advanced and incurable. Cochran reported that only 1 of 71 patients whose history and physical examination led to the diagnosis of recurrence was later cured surgically [20]. In the follow-up of rectal cancer patients, digital rectal exam, palpation of groin nodes, and bimanual pelvic exams in women can be useful. Laboratory evaluation The liver is the most common site of metastasis from colorectal carcinoma, representing approximately 50% of all recurrences. Often evaluation of liver function will be instituted. In a study of 327 patients with a history of colorectal cancer, 43 of the 56 patients subsequently found to have liver metastases also had elevated levels of alkaline phosphatase, for a sensitivity of 77% [21]. There were, however, 110 patients without metastases who also had elevated levels, for a false-positive rate of 34%. Thirteen patients with metastatic disease had normal levels for a false-negative rate of 4%. Given the low specificity, serum alkaline phosphatase levels have fallen out of favor as a screening tool. Tumor markers have been a controversial issue since their discovery. CEA is a glycoprotein oncofetal tumor-associated antigen whose biological function remains unclear. CEA has some value as a tumor marker in

1098

W.E. Longo, F.E. Johnson / Surg Clin N Am 82 (2002) 1091–1108

patients with known cancer. It has been shown to correlate with tumor volume [22], response to antitumor therapy [23], and likelihood of persistent tumor remaining after primary resection [24]. After curative resection, an elevated CEA should fall to normal within four to eight weeks. If this does not occur, then incomplete resection is likely. Unfortunately, 20% to 30% of individuals with recurrent disease will have normal CEA levels. The sensitivity and specificity of CEA in detecting recurrent colorectal cancer are approximately 70% and 80%, respectively [25]. The utility of rising CEA levels in predicting recurrent disease has also been confirmed in individuals with normal preoperative levels. Several studies have reported that CEA monitoring can lead to a diagnosis of recurrent disease before history and physical exam, with a median lead time of approximately six months [26–28]. A common question is what CEA level should trigger further work-up for recurrent disease or lead to exploratory surgery. Some use a single value above a certain cutoff point, such as 5 ng/ml. Others require two consecutive elevations, and still others require a specific slope of the rise in CEA levels to indicate likely recurrence [29]. At present there is no consensus, although most agree that a single test is less reliable than two or more. Northover et al randomized patients who had undergone curative surgery to an active intervention group or a control group [30]. CEA was measured in all patients at frequent intervals and, in the active intervention group, a rising CEA prompted further investigation, including second-look laparotomy when appropriate. Preliminary analysis showed no difference in survival between the two groups. Endoscopic evaluation Colonoscopy is an important part of follow-up for colorectal cancer patients, as they are at increased risk for the development of metachronous colorectal neoplasms. When possible, colorectal cancer patients should also undergo colonoscopy to the cecum preoperatively in order to identify synchronous lesions. Three to 7% of patients about to undergo colorectal cancer resection are found to have a second or synchronous malignancy at another location within the colon or rectum requiring resection. An additional 25% have adenomas that also require removal. Those individuals who present with obstruction or perforation should undergo colonoscopy within three to six months following resection. Endoscopy is useful in identifying metachronous lesions and anastomotic recurrences as well as synchronous lesions. Patients with prior rectal cancer are clearly at high risk for a second colorectal malignancy. Endoscopy can also be useful in the identification of intraluminal anastomotic recurrences. Anastomotic and local recurrence occurs more frequently with rectal cancers than with colon cancers. The great majority of those presenting with anastomotic recurrence are found to have either extensive local or systemic disease. In a retrospective analysis of anastomotic recurrences in

W.E. Longo, F.E. Johnson / Surg Clin N Am 82 (2002) 1091–1108

1099

50 patients, Rodriguez-Bigas found five patients alive without disease after five years [31]. Eighty-two percent of these anastomotic recurrences were identified on endoscopy. All of these patients were noted to have been asymptomatic at the time of diagnosis, and the recurrence was limited to the anastomosis on postoperative microscopic inspection. This study suggests that routine endoscopy, particularly in those with history of rectal cancer resection (low anterior resection), may lead to potential cure of recurrent disease. The ideal interval between endoscopic examinations has not been determined. The determination of what subset of patients benefit most has not been shown. A prospective trial is needed. Diagnostic imaging Plain films The value of plain films should not be underestimated. A flat plate of the abdomen may suggest a malignant obstruction or a mass effect from recurrent tumor. Extremity pain may prompt a radiograph revealing a pathologic fracture. Chest radiograph deserves discussion—between 5% and 10% of all patients undergoing operations for colorectal malignancy will develop pulmonary metastases, and the results of resection of pulmonary metastases have been encouraging. However, the optimal method and timing for detection of pulmonary metastases remains unclear. Graffner et al reported that recurrent disease was found on routine chest radiography in 3 of 47 patients, all of whom were asymptomatic [32]. Plain chest radiography has been the primary imaging modality described in the literature. If suspicion arises from the radiograph or from clinical findings, chest CT has a higher sensitivity and specificity than plain chest radiography. The ideal interval for postoperative chest radiography has not been determined. Rectal cancer has a greater propensity than colon cancer to metastasize to bone, with the vertebral column, skull, pelvic bones, and long bones of the extremity the most common sites. It is commonly felt that, in the absence of symptoms, routine surveillance of the bony skeleton in search of metastases is not warranted. If symptoms do develop, bone scanning is the most sensitive study for detection of such metastases. Contrast studies Local recurrence is recurrence at or adjacent to the site of the primary tumor, including the anastomosis, tumor bed, and regional lymph nodes. Barium enema has likely been the most used method of postoperative surveillance. Although it is not as sensitive as endoscopy (97%), air-contrast barium enema may reveal features such as an intraluminal filling defect, eccentric anastomotic narrowing, and local extrinsic mass effect. Ultrasound Ultrasound is a relatively inexpensive but quite operator-dependent modality of detecting liver metastases. It can be greatly influenced by

1100

W.E. Longo, F.E. Johnson / Surg Clin N Am 82 (2002) 1091–1108

patient-specific factors such as bowel gas overlying the liver, and obesity. The sensitivity of ultrasound in detecting metastatic liver disease is low (57%) and for lesions less than 1 cm in diameter it falls to around 20%. Intraoperative ultrasound, on the other hand, is very sensitive, and many believe it should be mandatory prior to resection. In a multi-institutional trial evaluating liver metastases from colorectal cancer, more than half of those determined to be resectable based on preoperative evaluation were later found at surgery to be unresectable by ultrasound [33]. Transrectal ultrasound is sometimes useful to diagnose recurrent rectal cancer. Ultrasound, like CT, is unable to distinguish malignant from benign masses. Additionally, transrectal ultrasound can only be used after sphincterpreserving surgery. Transvaginal ultrasound, however, is a possible alternative in women following abdominoperineal resection. Computed tomography CT scan has become the preferred method for evaluating local recurrence and has a sensitivity of approximately 95% [34]. In rectal cancer recurrences, it is important that rectal contrast be administered. CT scan can show not only the anastomosis, but the tumor bed, as well as regional lymph nodes. Any pelvic lymphadenopathy is likely secondary to recurrent disease. In the early postoperative period, a mass of soft-tissue density is not unexpected, due to the presence of granulation tissue, hemorrhage, edema, or fibrosis. Postradiation changes can produce streaky densities, presacral masses, or diffuse thickening of the rectal wall. All of these changes can easily be confused with recurrent disease. A soft-tissue mass may persist in the rectal bed for up to 24 months [35]. A CT at three to four months will frequently show a soft-tissue mass, even in patients who have no recurrence. A repeat study at 6 to 12 months often reveals a decrease in the size of the mass as well as an increase in its definition. Following its introduction, computed tomography soon became the gold standard for detecting hepatic recurrences and remains the primary modality used. Liver metastases from colorectal cancer appear as areas of low attenuation on noncontrast scans. Following a bolus of contrast material, these metastases often show early rim enhancement or hyperdensity, then an isodense period, and subsequently a hypodense phase. The appearance of the tumor on CT scan is therefore dependent on the timing of the intravenous contrast. CT with arterial portography involves repeat scanning four hours after the injection of contrast, and has been reported to further increase the sensitivity of CT scan for hepatic recurrence. The sensitivity of CT scan for metastatic colorectal carcinoma in the liver is high (78% to 90%) [36]. Magnetic resonance imaging For a time it was felt that MRI could allow for the differentiation of malignant from benign postoperative pelvic masses based on differences in

W.E. Longo, F.E. Johnson / Surg Clin N Am 82 (2002) 1091–1108

1101

signal intensity, but this has been discounted. Both CT and MRI are sensitive modalities for detecting masses, but neither is able to reliably distinguish malignant from benign lesions. Magnetic resonance imaging can be used to detect hepatic recurrences but current techniques provide for no better imaging than contrast enhanced CT and at a greater cost [33]. The sensitivity of both CT and MRI increases with increasing tumor size. Nuclear medicine Another methodology for the detection of recurrent disease is immunoscintigraphy A study of 42 patients suspected of having recurrent colorectal carcinoma was performed in order to compare conventional diagnostic methods (barium enema, CT) with immunoscintigraphy [37]. Eighty-three percent of the patients had recurrent disease localized by conventional methods and 57% had recurrent disease localized by immunoscintigraphy. Immunoscintigraphy has subsequently been found to be useful under specific circumstances: in the localization of recurrent disease when CEA levels are elevated but no lesion can be identified by other means (physical examination, barium enema, CT); and in the characterization of suspicious masses as malignant rather than undetermined. The reported sensitivity in detecting recurrent colorectal malignancy has ranged from 18% to 90%, and specificity has ranged from 76% to 97% [38–40]. Immunoscintigraphy tends to perform poorly for intrahepatic recurrences. Although immunoscintigraphy can be very useful in follow-up, it also has a notable false negative rate. Positron emission tomography Finally, the latest modality introduced and currently gaining great acceptance is positron emission tomography (PET). It is a functional imaging modality that relies on physiological and metabolic changes within tissue to allow for the detection of disease. (18F)-fluoro-2-deoxy-D-glucose (FDG) is a radioactive glucose analog accumulation of FDG within cells proportional to the metabolism and transport of glucose by these cells. PET scanning depends on the increased utilization of glucose by malignant cells to allow for detection. PET does not depend on gross morphological abnormalities and thus can be helpful in assessing patients with suspicion for recurrence based on elevated CEA levels or on physical or radiological findings that are indeterminate. Flanagan et al studied 22 patients with abnormal CEA levels and normal results of conventional methods of tumor detection [41]. Each patient underwent FDG-PET scanning and the results were compared with topathologic findings (n ¼ 9) and long-term radiologic and clinical follow-up. PET scanning was abnormal in 17 and normal in 5. Of 17 patients with abnormal PET scan results, 7 underwent resection for cure, 8 were later found to have extensive disease, and 2 were regarded as false positives. The 5 patients with negative PET scans were found to have no recurrence. Overall, the positive predictive value of PET scan imaging in

1102

W.E. Longo, F.E. Johnson / Surg Clin N Am 82 (2002) 1091–1108

these clinically difficult situations was 89% and the negative-predictive value was 100%. Multiple similar studies have confirmed the utility of FDG-PET in follow-up of selected colorectal cancer patients [42]. The relatively high cost and paucity of PET imaging centers have caused PET imaging to be limited primarily to university medical centers. More studies are needed to determine if immunoscintigraphy or PET imaging protocols in follow-up programs might affect survival and cure rates. A recent study suggested that FDG-PET scanning can predict those patients who would likely benefit from a laparotomy. In this study, it was determined that patients found to have a single site of otherwise resectable disease on standard imaging studies, an FDG- PET scan should be obtained to rule out additional sites of disease. For those patients who are found not to have additional sites of disease on FDG-PET scan, surgical resection is recommended, as it may prove curative. For those patients found to have additional sites of disease on FDGPET scan, laparotomy will more than likely reveal incurable disease [43]. Recurrence following curative resection for rectal cancer There are certain follow-up issues that are unique to rectal cancer: first, the fact that the pelvis represents the most common site for recurrence; and second, that local full-thickness excision represents a unique way to treat rectal cancer but not colon cancer. It is of paramount importance that accurate staging be performed before embarking on treatment of rectal cancer. Accurate staging will direct physicians in determining whether local therapy, radical surgery, or adjuvant therapy—either in the preoperative setting or postoperatively—should be employed. These efforts are obviously made to decrease the local recurrence rate and theoretically impact on the development of distant metastases. Accurate staging can also impact on the functional result following the appropriate procedure, giving the best oncological outcome. Patterns of recurrence are different in colon cancer than in rectal cancer. For example, follow-up programs after rectal cancer surgery should emphasize detection of locoregional recurrence in the pelvis and pulmonary metastases, whereas with colon cancer detection of liver metastases is more pressing. Furthermore, a variety of surgical treatments are available for the treatment of rectal cancer, including local excision, restorative proctosigmoidectomy, and abdominoperineal resection. Each of these procedures predisposes the patient to different forms of local recurrence. Following restorative procedures, a change in caliber of stool, rectal bleeding, or pelvic pain may represent either a luminal recurrence at the previous colorectal or coloanal anastomosis, or a recurrence in the pelvis. These symptoms should lead to prompt evaluation. Recurrence following abdominoperineal resection typically causes perineal pain or a perineal mass. In this setting, there is often a delay from the onset of symptoms to CT detection and diagnosis of recurrence.

W.E. Longo, F.E. Johnson / Surg Clin N Am 82 (2002) 1091–1108

1103

The reported recurrence rate following local excision ranges between 1% and 50%. Risk factors for local recurrence include transmural tumors, unfavorable histologic characteristics, and positive excision margins. Efforts to reduce the local recurrence rates have employed the addition of chemoradiation to local excision. Currently there is no consensus as to the most effective way to follow these patients. Often an anecdotal approach employing digital examination or direct inspection of the rectal mucosa is employed. Endorectal and endovaginal ultrasound have been used to detect asymptomatic local recurrence after rectal surgery. Among 120 patients followed in such a manner after radical restorative procedures, recurrences were detected in 17 (14%) [44]. There were two false positives. Currently, data are insufficient to support routine use of intrarectal ultrasound in patients who have undergone radical surgery for rectal cancer. Results of surveillance strategies Given the lack of strong statistical evidence for benefit, and the limitations of cost weighed against the numerous reasons for postoperative surveillance, a thorough yet cost-effective and minimally invasive strategy is needed. Until 1995, there were no randomized clinical trials published on follow-up. In 1994, a meta-analysis of seven nonrandomized trials involving over 3000 patients and comparing routine and intensive follow-up determined that, although more asymptomatic recurrences were detected and resected in the intensive follow-up group, there was no significant difference in survival [45]. When the analysis was restricted to those studies that included CEA measurements, there was an apparent 9% increase in five-year survival in the intensively investigated group. A previous survey of the members of the American Society of Colon and Rectal Surgeons (ASCRS) examined patterns of surveillance in patients whose colon cancer had been treated with potentially curative surgery [46]. This survey assessed the use of nine follow-up measures, including office visits, complete blood counts, liver function tests, carcinoembryonic antigen (CEA), chest radiograph, bone scan, computed tomography (CT), colonoscopy, and flexible sigmoidoscopy. Among the responding members of this organization of colorectal cancer experts, a wide variation in practices was found, and no clear consensus pattern could be identified. In a similar study of the Society of Surgical Oncology (SSO) there was also wide variability in test-ordering patterns among respondents. The practices of SSO members differed moderately from those of the previously surveyed ASCRS members. Interestingly, the charge differential between high- and low-intensity follow-up regimens for each patient cohort was about $800 million per year [47]. Charges vary extensively across follow-up strategies, with no indication that higher cost strategies increase survival or quality of life. It can be presumed that the results of these studies and the similar lack of consensus would pertain to rectal cancer as well.

1104

W.E. Longo, F.E. Johnson / Surg Clin N Am 82 (2002) 1091–1108

A difficult continuing issue concerns the effect of follow-up on survival. A recent study from Edinburgh sought to determine whether frequent liver imaging could detect liver metastases suitable for surgical or chemotherapeutic intervention at an asymptomatic stage [48]. Among 157 patients, 21 out of 24 asymptomatic patients developed histologically proven liver metastases. Twelve patients had metastases diagnosed by both ultrasound and CT scanning and the remaining 9 had metastases that were diagnosed by CT alone. Patients with liver metastases detected at an asymptomatic stage had a median survival of 16 months (range 7–41 months) from the time of diagnosis, whereas survival in the symptomatic group was less than four months. In randomized trials [49,50] comparing minimal and intensive follow-up, recurrence rates were similar in both groups, but the tumor recurrence in the intensive group were detected on an average nine months earlier, often at an asymptomatic stage. There was, however, no difference in overall or cancerrelated survival between the two groups. A similar result was reported by Northover [30] where a rising CEA prompted further investigation. In 1999, in an effort to determine the most effective, evidence-based, postoperative surveillance strategy for the detection of recurrent colon and rectal cancer, The American Society of Clinical Oncology (ASCO) published their recommendations in the Journal of Clinical Oncology [51]. This was based on a complete MEDLINE search performed of the past 20 years of medical literature. All tests described in the literature for postoperative monitoring were considered and an expert panel recommended a postoperative monitoring schema, which is summarized in the box below. Patients with rectal cancer, specifically those with stage II or stage III disease who did not receive pelvic radiation, should have direct imaging (flexible proctosigmoidoscopy) of the rectum at periodic intervals. Routine complete blood count (CBC), liver function tests, chest radiograph, CT scanning, and other pelvic imaging are not recommended unless directed by symptoms or elevated CEA.

Postoperative serum CEA testing should be performed every two to three months in patients with stage II or III disease for ‡2 years after diagnosis. • A clinical history and pertinent physical examination should be performed every three to six months for the first three years and annually thereafter. • Colonoscopy should be performed every three to five years for patients with previous colon or rectal cancers to detect new cancers or polyps. •

W.E. Longo, F.E. Johnson / Surg Clin N Am 82 (2002) 1091–1108

1105

Table 1 Results of selected series regarding surgical outcome of resected recurrences Author/Reference

Year

No. recurrences

No. intended curative operations

5 year survival (%)

Ohlsson Makela [50] Audisio Peethambaram Graham Goldberg [18] Bleeker [52]

1995 1995 1996 1997 1998 1998 2001

35/107 (32%) 22/52 (42%) 141/505 (28%) 97/312 (31%) 421/1356 (31%) 548/1247 (44%) 213/496 (43%)

8/35 (23%) 22/22 (100%) 32/141 (23%) 16/97 (16%) 96/421 (23%) 222/548 (41%) 42/213 (19.7%)

67% 57% NR NR NR 23% 40%

NR, Not reported.

Recently, Bleeker examined the cost of follow-up after adjuvant treatment of patients with Dukes C colonic cancer. Among a cohort of 496 patients treated with chemotherapy over a one-year period, the mean cost of diagnostic procedures per curative resected recurrence for patients amenable to salvage therapy was $9011. Potentially curable recurrences were detected primarily by liver imaging and colonoscopy. The yield of CEA measurement, chest radiography, and physical examination was relatively low and expensive [52]. A concerning issue is that socioeconomic differences in the receipt of colorectal cancer surveillance care following treatment with curative intent have been demonstrated [53]. The results of selected series regarding surgical outcome of resected recurrences are depicted in Table 1. Summary Many advances have been made in the field of colorectal cancer followup since the pioneering efforts of Wangensteen and others with second-look operations in the 1950s. The understanding of the biology and natural history of colorectal malignancy has been advanced. Diagnostic methods for detection of recurrent disease have also advanced tremendously with CEA monitoring, immunoscintigraphy, CT, MRI, and PET imaging. As has been discussed in this article, however, no strategy of postoperative follow-up has been shown unequivocally to produce improved survival benefit or cure rate. It is quite possible that benefit will be shown, but well-controlled trials will be required. Cost considerations will likely prove important, because the rate of detection of curable disease will likely be low. Quality of life issues will also be important in such trials. Better treatment and outcome of recurrent disease would provide a strong rationale for vigorous postoperative surveillance. New recommendations are currently evolving [54]. Early diagnosis seems likely to enhance the curability of both local and distant relapses and second primary tumors. Furthermore, there may be a survival and quality of life advantage that results from the early institution of chemotherapy, even for those tumors found to be inoperable [55]. In devising a plan for

1106

W.E. Longo, F.E. Johnson / Surg Clin N Am 82 (2002) 1091–1108

follow-up in patients, it is important to recognize the anatomic and temporal patterns of recurrence as well as their relationships to the initial tumor staging. Although there is little proof that the identification of recurrent disease in follow-up programs increases the likelihood of resectability, cure, or prolonged survival, many physicians have witnessed successful treatment of recurrent colorectal cancer. These anecdotal experiences, the unproven belief that follow-up is beneficial, and traditions imparted during training are among the likely motivating factors for most physicians caring for colorectal cancer patients. References [1] Anderson JH, Hole D, McArdle CS. Elective versus emergency surgery for patients with colorectal cancer. Br J Surg 1992;79:701–9. [2] Khuri SF, Daley J, Henderson WG, et al. The National Veterans Administration Surgical Risk Study: risk adjustment for the comparative assessment of the quality of surgical care. J Am Coll Surg 1995;180:519–31. [3] Northover JA. Follow-up after colorectal cancer resection: is it worthwhile? In: Scholefield JH, editor. Challenges in colorectal cancer. London: Blackwell Science Ltd.; 2000. p. 63–80. [4] Lavery IC. Preoperative Evaluation. In: Cohen AM, Winawer SJ, editors. Cancer of the colon, rectum and anus. New York: McGraw-Hill, Inc.; 1995. p. 425–29. [5] Kerner BA, Oliver GC, Eisenstadt TE. Is preoperative computerized tomography useful in assessing patients with colorectal carcinoma? Dis Colon Rectum 1993;36:1050–3. [6] Saclarides TJ. Endorectal ultrasound. Surg Clin North Am 1998;78:237–49. [7] Mukai M, Sadahiro S, Yasuda S, et al. Preoperative evaluation by whole-body 18Fflurodeoxyglucose positron emission tomography in patients with primary colorectal cancer. Oncol Rep 2000;7:85–7. [8] Wiggers T, Arends JW, Volovics A. Regression analysis of prognostic factors in colorectal cancer after curative resections. Dis Colon Rectum 1988;31:33–41. [9] Philips RKS, Hittinger R, Blesovsky L, et al. Local recurrence following curative surgery for large bowel cancer. The overall picture. Br J Surg 1984;71:12–6. [10] Longo WE, Virgo KS, Johnson FE, et al. Risk factors for morbidity and mortality after colectomy for colon cancer. Dis Colon Rectum 2000;43:83–91. [11] Longo WE, Virgo KS, Johnson FE, et al. Outcome after proctectomy for rectal cancer in Department of Veterans Affairs hospitals. Ann Surg 1998;228:64–70. [12] Macintosh EL, Rodriguez-Bigas AR, Petrelli NJ. Colorectal carcinoma. In: Johnson FE, Virgo KS, editors. Cancer patient follow-up, St. Louis: Mosby; 1997. p. 118–31. [13] Safi F, Link KH, Beger HG. Is follow-up of colorectal cancer patients worthwhile? Dis Colon Rectum 1993;36:636–44. [14] Bohm B, Schwenk W, Hoche HP, et al. Does methodological long-term follow-up affect survival after curative resection of colorectal carcinoma? Dis Colon Rectum 1993;36:280–6. [15] Ohlsson B, Bleland V, Ekberg H, et al. Follow-up after curative surgery for colorectal carcinoma: randomized comparison with no follow-up. Dis Colon Rectum 1995;38:619–26. [16] Steele G. Standard postoperative monitoring of patients after primary resection of colon and rectum cancer. Cancer 1993;71:4225–35. [17] Holm A, Bradley G, Aldrete JS. Hepatic resection of metastases from colorectal carcinoma. Ann Surg 1989;209:428–34. [18] Goldberg RM, Fleming TR, Tangen CM, et al. Surgery for recurrent colon cancer: strategies for identifying resectable recurrence and success rates after resection. Ann Intern Med 1998;129:27–35.

W.E. Longo, F.E. Johnson / Surg Clin N Am 82 (2002) 1091–1108

1107

[19] Beart RW, O’Connell MJ. Postoperative follow-up of patients with carcinoma of the colon. Mayo Clin Proc 1983;58:361–3. [20] Cochran JP, Williams JT, Faber RG, et al. Value of outpatient follow-up after curative surgery for carcinoma of the large bowel. BMJ 1980;280:593–5. [21] Baden H, Anderson B, Augustenborg G. Diagnostic value of gamma-glutamyl transpeptidase and alkaline phosphatase in liver metastases. Surg Gynecol Obstet 1971;133: 769–73. [22] Bronstein RR, Steele G, Ensminger W. The use and limitations of serial plasma carcinoembryonic antigen (CEA) levels as a monitor of changing metastatic liver tumor volume in patients receiving chemotherapy. Cancer 1980;46:266–72. [23] Mayer R, Garnick M, Steele G. CEA as a monitor of chemotherapy in disseminated colorectal cancer. Cancer 1978;42:1428–32. [24] Goslin R, Steele G, Maclntyre J. The use of preoperative plasma CEA levels for the stratification of patients after curative resection of colorectal cancers. Ann Surg 1980;192: 747–51. [25] Hine FR, Dyke PW. Serum CEA testing in the postoperative surveillance of colorectal carcinoma. Br J Surg 1984;49:689–93. [26] Armitage NC, Davidson A, Tsikos D, et al. A study of the reliability of carcinoembryonic antigen blood levels in following the course of colorectal cancer. Clin Oncol 1984;10:141–7. [27] McCall JL, Black RB, Rich CA, et al. The valve of serum carcinoembryonic antigen in predicting recurrent disease following curative resection of colorectal cancer. Dis Colon Rectum 1994;37:875–81. [28] Staab HJ, Anderer FA, Stumpf E, et al. Eight-four potential second-look operations based on sequential carcinoembryonic antigen determinations and clinical investigations in patients with recurrent gastrointestinal cancer. Am J Surg 1985;149:198–204. [29] Carl J, Bentzen SM, Norgaard-Pedersen B, et al. Modelling of serial carcinoembryonic antigen changes in colorectal cancer. Scan J Clin Lab Invest 1993;53:751–5. [30] Northover J, Houghton J, Lennon T. CEA to detect recurrence of colon cancer. JAMA 1994;272:31. [31] Rodriguez-Bigas MA, Stulc JP, Davidson B, et al. Prognostic significance of anastomotic recurrence from colorectal adenocarcinoma. Dis Colon Rectum 1992;35:838–42. [32] Graffner H, Holtberg B, Johansson B, et al. Detection of recurrent cancer of the colon and rectum. J Surg Oncol 1985;28:156–9. [33] Steele G, Bleday R, Meyer RJ, et al. A prospective evaluation of hepatic resection for colorectal carcinoma metastases to the liver: Gastrointestinal Tumor Study Group protocol 6584. J Clin Oncol 1991;9:1105–12. [34] Moss AA, Thoeni RF, Schnyder P. Value of computed tomography in the detection and staging of recurrent rectal carcinomas. J Comput Assist Tomogr 1981;5:870–34. [35] Kelvin FM, Korobkin M, Heaston DK. The pelvis after surgery for rectal carcinoma: serial CT observations with emphasis on non-neoplastic features. AJR 1983;141:959–64. [36] Ward BA, Miller DL, Frank JA. Prospective evaluation of hepatic imaging studies in the detection of colorectal metastases: Correlation with surgical findings. Surgery 1989;105:180–7. [37] Holting T, Schlag P, Steinbacher M, et al. The value of immunoscintigraphy for the operative retreatment of colorectal cancer. Cancer 1989;64:830–3. [38] Goldenberg DM, Larson SM. Radioimmunodetection in cancer identification. J Nucl Med 1992;33:803–14. [39] Abdel-Nabi H, Doerr RJ. Clinical applications of indium-111-labeled monoclonal antibody imaging in colorectal cancer patients. Semin Nucl Med 1993;23:99–113. [40] Rutgers EJT. Radioimmunotargeting in colorectal carcinoma. Eur J Cancer 1995;31A: 1243–7. [41] Flanagan FL, Dehdashti F, Ogunbigi OA, et al. Utility of FDG-PET for investigating unexplained plasma CEA elevation in patients with colorectal cancer. Ann Surg 1998;227: 319–23.

1108

W.E. Longo, F.E. Johnson / Surg Clin N Am 82 (2002) 1091–1108

[42] Keogan MT, Lowe VJ, Baker ME, et al. Local recurrence of rectal cancer: evaluation with F-18 fluorodeoxyglucose PET imaging. Abdom Imaging 1997;22:332–7. [43] Libutti SK, Alexander HR, Choyke P, et al. A prospective study of 2-18F fluoro-2-deoxyD-glucose/positron emission tomography scan, 99mTc-labelled arcitumomab (CEA scan), and blind second-look laparotomy for detecting colon cancer recurrence in patients with increasing carcinoembryonic antigen levels. Ann Surg Oncol 2001;8:779–86. [44] Masgagni D, Corbellini L, Urciuoloi P, et al. Endoluminal ultrasound for early detection of local recurrence of rectal cancer. Br J Surg 1989;76:1176–80. [45] Bruinvels DJ, Stiggelbout AM, Kievut J, et al. Follow-up of patients with colorectal cancer: a meta-analysis. Ann Surg 1994;219:174–82. [46] Vernava AM, Longo WE, Virgo KS, et al. Current follow-up strategies after resection of colon cancer: results of a survey of members of the American Society of Colon and Rectal Surgeons. Dis Colon Rectum 1994;37:573–83. [47] Virgo KS, Wade TP, Longo WE, et al. Surveillance after curative colon cancer resection: practice patterns of surgical specialists. Ann Surg Oncol 1995;2:472–82. [48] Howell JD, Wotherspoon H, Leen E, et al. Evaluation of a follow-up programme after curative resection for colorectal cancer. Br J Cancer 1999;79:308–10. [49] Kjeldsen B, Kronberg O, Fenger C, et al. A prospective randomized study of follow-up after radical surgery for colorectal cancer. Br J Surg 1997;84:666–9. [50] Makela JT, Laitnen SO, Kairaluoma MI. Five year follow-up after radical surgery for colorectal cancer: results of a prospective randomized trial. Arch Surg 1995;130:1062–7. [51] Desch CE, Benson AB, Smith TJ, et al. Recommended colorectal cancer surveillance guidelines by the American Society of Clinical Oncology. J Clin Oncol 1999;17:1312–21. [52] Bleeker WA, Mulder NH, Hermans J, et al. Value and cost of follow-up after treatment of patients with Dukes C colonic cancer. Br J Surg 2001;88:101–6. [53] Lafata JE, Johnson CC, Ben-Menachem T, et al. Sociodemographic differences in the receipt of colorectal cancer surveillance care following treatment with curative intent. Med Care 2001;39:361–72. [54] Bast RC, 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 2001;19:1865–78. [55] O’Dwyer PJ, Stevenson JP, Haller DG, et al. Follow-up pf Stage B and C colorectal cancer in the United States and France. Semin Oncol 2001;28:45–9.