Key points

Eighty percent of patients with colorectal cancer (CRC) present with local or regional disease. For these patients, the general plan of treatment is clear: surgery with the intent of cure. However, about 20% of newly diagnosed patients continue to present with synchronously diagnosed stage 4 disease. For these patients, the treatment plan is less obvious. Despite their advanced stage of disease, a subset of stage 4 patients are potentially curable. This subset is of patients with so-called oligometastasis. The rest of these patients have truly disseminated disease and cannot be cured. Some patients with disseminated disease declare their condition immediately. Others patients with disseminated disease may at first masquerade as potentially curable metastatic disease, only to blossom into full-blown disseminated disease later in their course. The current staging system does not address these subsets.

The problem the surgeon faces in treating these stage 4 CRC patients is to try to identify the burden of metastatic disease and to provide an appropriate treatment plan. Patients with minimal metastatic disease that is potentially treatable for cure or long-term survival should be offered aggressive treatment. The approach is usually resection of both the primary and the metastatic tumor. Conversely, patients with disseminated disease should not undergo surgery if it offers no chance of cure or meaningful palliation.

When faced with synchronous stage 4 disease, the surgeon must develop a treatment plan that addresses whether to proceed directly to surgery, and if so, what operation or operations should be performed and in what order; whether to recommend chemotherapy and/or radiation therapy first and then re-evaluate; or whether to avoid surgery altogether in favor of long-term chemotherapy or a hospice approach.

Because there are so many different types and degrees of metastatic disease, and because the primary bowel tumors present in various ways, the decision-making process for these patients can be complex. These patients are not easily categorized and their treatment plans defy development of precise algorithms. Treatment of these patients requires more judgment than perhaps any other form of cancer. Multidisciplinary input is helpful. This input includes experts in medical and radiation oncology as well as diagnostic and interventional radiology. Input from surgeons with expertise in surgery of the large bowel, liver, lung, and occasionally other organs may be required.

One approach to planning treatment of a new patient with colon and rectal cancer is to answer a series of questions:

• •
  

  Does the patient have synchronous metastatic disease?

  
  
• •
  

  Is immediate surgery required for palliation of the primary tumor?

  
  
• •
  

  What is the type and extent of metastatic disease?

  
  
• •
  

  Is the metastatic disease amenable to potentially curative surgery?

  
  
• •
  

  Should the metastatic disease be treated before the primary disease?

  
  
• •
  

  If the patient is not curable, is there an opportunity for meaningful prolongation of life by either surgical or nonsurgical measures?

Does the patient have synchronous stage 4 cancer?

The first suspicion of metastatic disease may arise during the history and physical examination. However, most metastatic disease is first identified by staging investigations. Computed tomographic (CT) scan of the chest, abdomen, and pelvis has several advantages. Good equipment is widely available and not operator dependent. Radiologists skilled in interpreting the study are also widely available. The study provides information for planning a resection and will generally confirm the location of the tumor.

CT scan is an excellent study for examining the liver, the most common site of distant metastasis. Ideally the study should include both arterial and portal phases of intravenous contrast. A noncontrast CT scan has poor sensitivity for detecting small liver lesions. Occasionally contrast will be omitted because of contrast allergy or simply deferred based on patient preference or issues with intravenous access. In these cases the surgeon should decide how critical it is to delay surgery in favor of better imaging. Options are to repeat the CT study of the liver with contrast, or to consider a complementary study such as magnetic resonance imaging or ultrasound.

Magnetic resonance imaging with contrast is comparable to or better than CT scan with some advantages, such as better specificity for benign liver lesions and more precise demarcation of intrahepatic vascular anatomy. It is also superior for evaluating response to preresection chemotherapy. Because this test is longer, more uncomfortable, and more expensive, it is usually not ordered routinely. Ultrasound in highly skilled hands is also of some value, but this is highly operator dependent and, because the cross-sectional imaging studies have gained greater sophistication, ultrasound has become a less important tool. Its main use now is as an intraoperative study, performed by the surgeon during hepatic resection.

In the past a standard posteroanterior and lateral chest radiograph was considered to be an adequate preliminary tool for identification of lung metastases. More recently, inclusion of chest CT has gained favor. Many of the small, nonspecific parenchymal lesions that show up on chest CT turn out to be of no clinical significance, creating distractions for the medical team and anxiety for the patient. These small nonspecific CT scan lesions in a patient with a normal chest radiograph are often benign and should not play a major role in operative planning.

Positron emission tomography should not be a routine initial staging tool for CRC patients. At this time, positron emission tomography should mainly be used to help define equivocal CT scan findings, and before aggressive surgery for metastatic disease, such as major hepatic or pulmonary resection.

Does the patient have synchronous stage 4 cancer?

The first suspicion of metastatic disease may arise during the history and physical examination. However, most metastatic disease is first identified by staging investigations. Computed tomographic (CT) scan of the chest, abdomen, and pelvis has several advantages. Good equipment is widely available and not operator dependent. Radiologists skilled in interpreting the study are also widely available. The study provides information for planning a resection and will generally confirm the location of the tumor.

CT scan is an excellent study for examining the liver, the most common site of distant metastasis. Ideally the study should include both arterial and portal phases of intravenous contrast. A noncontrast CT scan has poor sensitivity for detecting small liver lesions. Occasionally contrast will be omitted because of contrast allergy or simply deferred based on patient preference or issues with intravenous access. In these cases the surgeon should decide how critical it is to delay surgery in favor of better imaging. Options are to repeat the CT study of the liver with contrast, or to consider a complementary study such as magnetic resonance imaging or ultrasound.

Magnetic resonance imaging with contrast is comparable to or better than CT scan with some advantages, such as better specificity for benign liver lesions and more precise demarcation of intrahepatic vascular anatomy. It is also superior for evaluating response to preresection chemotherapy. Because this test is longer, more uncomfortable, and more expensive, it is usually not ordered routinely. Ultrasound in highly skilled hands is also of some value, but this is highly operator dependent and, because the cross-sectional imaging studies have gained greater sophistication, ultrasound has become a less important tool. Its main use now is as an intraoperative study, performed by the surgeon during hepatic resection.

In the past a standard posteroanterior and lateral chest radiograph was considered to be an adequate preliminary tool for identification of lung metastases. More recently, inclusion of chest CT has gained favor. Many of the small, nonspecific parenchymal lesions that show up on chest CT turn out to be of no clinical significance, creating distractions for the medical team and anxiety for the patient. These small nonspecific CT scan lesions in a patient with a normal chest radiograph are often benign and should not play a major role in operative planning.

Positron emission tomography should not be a routine initial staging tool for CRC patients. At this time, positron emission tomography should mainly be used to help define equivocal CT scan findings, and before aggressive surgery for metastatic disease, such as major hepatic or pulmonary resection.

Is immediate surgery required for palliation of the primary tumor?

Although the presence of the synchronous metastasis is ominous, 3 tumor characteristics trump the immediate concerns about the distant tumor spread: obstruction, bleeding, and pain due to invasion of local structures.

In this clinical context, obstruction is the presence of cramps, abdominal distension, nausea, vomiting, constipation, or the passage of frequent loose stools. One should probably not include asymptomatic patients who are described as being “obstructed” by the endoscopist because of an inability to pass a colonoscope through the lesion. Some of these “endoscopic obstructions” include patients who may go for many months before developing symptoms, which is especially true of lesions proximal to the splenic flexure where the stool is liquid.

For patients with bona fide high-grade obstruction, the best course of action is usually operative. Although chemotherapy or radiation therapy may reduce the size of the primary tumor, the therapeutic benefits are at best slow to occur. In the truly obstructed patient with stage 4 CRC, there are 4 options: resection with primary anastomosis; resection with end stoma; proximal diversion or bypass without resection; or placement of an endoluminal stent or other endoluminal therapy. Ideally, the lesion can be resected, but the exact strategy will be determined by the patient’s general state of health.

On first impression, endoscopic placement of an endoluminal stent seems quite attractive. The advantages of this procedure are complete avoidance of a surgical incision or general anesthesia. However stenting has several limitations. The technique can be difficult and is not widely practiced. There is a risk of perforation, either during insertion or as a delayed complication. The luminal diameter of the stent is relatively small compared with the normal colon. When inserted into the left side of the colon, where stool is normally solid, the patient must be attentive to maintaining a liquid stool to prevent obstruction of the stent by solid stool. This maintenance has been compared with being on a perpetual bowel preparation for the patient. Finally, the stent cannot be easily used for low rectal cancers because of the tendency for the distal tip of the stent to create severe tenesmus or occasionally extrude through the anal canal. Stent migration can occur. Nevertheless stenting is appropriate in a small subset of patients. Other techniques such as laser fulguration are even less widely practiced.

Less commonly a large endophytic colon cancer may invade an adjacent loop of small bowel, causing a small bowel obstruction. This form of obstruction must be carefully differentiated from a colon obstruction. Obviously a proximal colostomy or endoluminal stent would do nothing to alleviate this obstruction, putting the patient through an unnecessary and ineffective surgical or endoluminal procedure. In these cases a direct surgical approach to the tumor is required.

In patients with heavily bleeding cancers, surgical resection may be the only way to prevent the need for frequent transfusions, even in incurable stage 4 disease. This bleeding can be particularly bothersome if the patient has a requirement for long-term anticoagulation with warfarin, heparin, clopidogrel, or any of the growing number of new potent anticoagulants. In this setting, simple diversion or bypass will likely not stop the bleeding.

When considering anemia as an indication for immediate surgery, the surgeon must consider the type of bleeding. Anemia is a common presenting finding in CRC patients, particularly with right colon cancers. Most of these patients present with an iron deficiency anemia that has developed insidiously. This anemia must be distinguished from the previously described pattern of acute lower gastrointestinal bleeding. Although initial correction of severe anemia may require transfusion, the need for ongoing transfusions may be obviated with daily oral iron replacement and urgent resection is generally not needed.

Pain caused by aggressive T4 invasion of adjacent structures by the tumor may be another indication for early resection in synchronous stage 4 CRC, including patients in whom a large tumor is invading the parietes, causing somatic pain. Before attempting resection, the surgeon should be confident that an R0 excision can be achieved. Leaving tumor behind will likely do nothing to alleviate the somatic pain. The classic example of an unresectable tumor would be sacral or deep pelvic side wall invasion by a rectal cancer. In these patients, pain usually indicates that the tumor is unresectable by conventional techniques.

If none of these urgent indications for surgery are present in the stage 4 patient, then the surgeon should consider whether the patient would be better served initially by a nonoperative approach. These decisions are probably best approached on a multidisciplinary level, ideally in the setting of a multidisciplinary cancer conference.