The multidisciplinary management of liver metastases in colorectal cancer

Introduction: Multidisciplinary team approach

The diagnosis, staging, and treatment of colorectal cancer (CRC), as well as the management of potential medical and psychological complications, requires a multidisciplinary or team approach. Throughout a patient’s cancer journey, the central tenet for the team is providing continuous optimal care from all aspects, of a typically incurable condition. Evidence-based medicine provides changing guidelines for the benefit, almost exclusively, of a single intervention in the context of controlled clinical trials, where patients fit very selective, often rigorous, criteria. However, the real-life challenges are significantly more complex and more commonplace as the whole colorectal cancer patient population encompasses a broader and more multifaceted spectrum and invariably need multiple interventions at various levels of evidence. In reality the impact of potential toxicities of sequential treatments and impact on quality of life are rarely disclosed within the published literature. The challenge is therefore for each patient to be analyzed systematically by the colorectal multidisciplinary team, and then to use a contemporaneous evidenced-based approach to justify and facilitate a sequence of clinical interventions. Patients should be made aware and be offered choices, where options are available, as there is often a limited impact of treatments on quantity of life and the potential effects on the quality of life. An independent patient-linked advocate, which is usually an “advanced nurse practitioner” (ANP) or a “clinical nurse specialist’ (CNS), helps immensely, in what is a terrifying situation for patients and their famili es.

The role of complex liver interventions and surgery for colorectal liver metastases is discussed later in this chapter and is one of the specific areas where the prospective planned management of colorectal cancer patients is changing rapidly. It may have a major, potentially curative, impact on some patients and may also contribute to prolonging survival in as many as 20% of patients presenting with, or relapsing with, metastatic disease.

Epidemiology and staging

Colorectal cancer is the second leading cause of cancer deaths in developed nations. There are approximately 1,000,000 cases per year globally with more than half a million deaths. There is a wide geographical global incidence range, with most Western countries at 20–40 per 100,000 population. Colorectal cancer is the third most common cancer in the United States, with more than 100,000 cases of colorectal cancer diagnosed. Colon cancer affects both sexes with approximately equal frequency; however, tumors in the rectum appear more commonly in men at a ratio of 1.5:1 to women. Colorectal cancer incidence is gradually decreasing in western countries. The probability of developing colorectal cancer rises sharply with age, with 83% of all cases arising in people who are 60 years or older and median age at diagnosis over 70 years. The lifetime risk for men of colorectal cancer is estimated to be 1 in 18, and 1 in 20 for women.

The overall survival (mortality) is stage-specific, with the 5-year overall mortality decreasing incrementally over the last 30 years. This is due to the synergy between individual advances in diagnostics, surgery, and oncology. The overall 5-year survival of all CRC cases in Western countries ranges from 52% to 65%. A total of 80%–85% of patients with colon cancer have some symptoms at the time of diagnosis. At disease presentation, around 55% of people have advanced cancer that has either spread to lymph nodes or other organs, or is so locally advanced that surgery is unlikely to be curative (stage III or IV or Dukes’ C or D). More advanced stages of local and distant spread correlate inversely with survival; however, positive anomalies have now appeared, because of improvements in advanced colorectal cancer management. The first example is with the advent of adjuvant chemotherapy, in that the prognostic outcomes of substages of II and III now overlap: Stage IIa and stage IIIa have similar survival rates, with stage IIb (T4N0) actually having significantly worse outcomes compared to stage IIIa.

The second example is more recent, and is almost certainly attributable to successful liver surgery for limited CRC metastases. The prognostic outcomes of substages of III and IV now also overlap in that the worse subsets of stage III and best of stage IV (limited surgically operable metastases) have similar survival rates. The staging systems are thus only tools in assessment but increasingly are understood to be one of many factors that influence outcomes. Increasingly, patient-related clinical factors and biolog parameters, including chemosensitivity and genetic profiles, often outweigh the impact of staging both adversely and favorably, in ways that we as yet cannot predicate.

Tissue diagnosis is mandatory in all cases, unless the patient is not fit for any treatment interventions. Modern centers now add immediate genetic testing for somatic mutations, as these have major implications for systemic treatment options, both from a predictive and prognostic point of view. At a minimum, KRAS and BRAF testing should be done at the outset, in all metastatic cases. Many centers have moved to multigene testing systems, as small subsets of cancers have “druggable” mutations with targeted therapies that are already available.

A strong family history or multiple recurrent polyps should invoke potential discussions about family, genetic, and mutation testing with qualified geneticists and counselors. A detailed discussion on this subject is beyond the scope of this chapter.

Staging is mandatory prior to discussion on treatment planning. A minimum of a modern high-quality computed tomography (CT) scan of chest/abdomen and pelvis should be performed in a patient with confirmed cancer. The focus of the radiology should be to look specifically at common sites and patterns of spread of colon cancer: (in descending order) liver, peritoneum, lung and distant lymph nodes. Note that in contrast to the other common adenocarcinomas (i.e., breast, lung, and prostate), bone, spinal cord and brain, and CRC metastases are uncommon and occur in less than 5% of metastatic patients at presentation.

Additional radiologic tests are extremely useful in rectal cancer for planning management. An MRI of the pelvis in rectal tumors is used to assess whether primary surgical curative removal is potentially feasible and whether any neoadjuvant chemoradiotherapy is warranted to try and improve local control and overall outcomes. Modern evidence suggests that MRI is better than CT for local staging of primary rectal cancer, particularly in assessing unresectable disease (or “threatened margins”) and identification of cancer-positive lymph nodes. There is, however, wide variability between results reported in different studies and, which highlights the need for dedicated and trained site-specific radiologists to minimize the false-negative and false-positive rates of determining involved lymph nodes. In a phase II prospective series (MERCURY) of 712 patients, the correlation between MRI and pathologic staging after surgery alone was 0.92, with an observed agreement of 84% (kappa 0.49, 95% CI: 0.35–0.61) between the MERCURY assessment of prognosis and pathology findings after primary surgery.

In situations where liver or lung surgery (metastatectomy) is contemplated, dual, triple, or even quadruple imaging modalities are often necessary. Liver MRI, liver ultrasonography (preferably with contrast enhancement with microbubbles) and F-fluorodeoxyglucose positron emission tomography ( F-FDG-PET) scanning are often complimentary to a CT. F-FDG-PET is particularly useful in detecting occult extrahepatic disease. Because of the difficulty in localizing structures with an F-FDG-PET, this investigation is always performed in conjunction with a separate CT contrast-enhanced scan, despite dual-modality combination advances (CT-PET, which do not have the same specificity or sensitivity for liver imaging).

Although blood tests are frequently nonspecific, they can impart useful information for treatment planning and prognostication. For example, a low serum albumin, high alkaline phosphatase, and high lactate dehydrogenase (LDH) level and possibly high neutrophil and platelet counts are poor prognostic factors. In advanced disease, the liver and renal function can significantly alter the choice or doses of chemotherapeutic agents or interventional therapy that can be offered safely.

Specific tumor markers for colorectal cancer have become commonplace through their clinical utility. About two-thirds of CRC are associated with an elevated serum CEA (carcino-embryonic antigen). This and the serum CA19.9 antigen can be very useful tools in monitoring treatment response in conjunction with other modalities such as radiology. In isolation, they have low specificity and so always need validation with imaging to confirm their trend. A crucial learning point is that tumor markers are not validated to confirm changes in treatment pathway decisions and are not allowable in regulatory clinical trials as valid end points of response or progression.

Systemic chemotherapy

In deciding treatment options, the primary question is whether the patient has a technically or potentially curable malignancy. The secondary and equally important point is whether it is clinically appropriate for the patient to be subject to this potentially curative approach.

The only clear level 1, curative modality is resection. All other modalities of treatment can add or accentuate the benefit of resection, but are rarely curative in their own right. Resection of the primary colonic or rectal tumor is the principal first-line treatment in the majority of colorectal cancer patients. However, in the context of advanced metastatic disease, it is noncurative and predominantly performed for palliation of symptoms alone. In more recent practice, and with the improvement of systemic chemotherapy tumor response rates, the frequency of the first intervention being primary colorectal surgery is decreasing and is as low as 60% now in contemporaneous clinical studies. When curative resection is not possible, patients may benefit from palliative resection, which is becoming increasingly minimally invasive, and more alternatives such as endoluminal stenting to relieve partial or subacute obstruction are now available as means to expedite the systemic treatment.

Indeed, a risk of attempting primary surgery first is significant delay in starting treatment in metastatic disease, which is the usual cause of constitutional deterioration in advanced patients and ultimately responsible for deaths.

High-quality surgery is thus crucial to patients’ survival, and the surgical outcomes are the platform of overall success of the MDT. Hence, specialist colorectal cancer surgeons should be able to demonstrate low tumor involvement at the margins of the excised cancer specimens, low rates of surgical complications, and high survival rates among their patients. If surgery is not the planned primary modality, a rediscussion at various time points, should ensue, as to whether there is an appropriate treatment window, and where it can be integrated without significant disturbance to the systemic objectives. There is mounting retrospective evidence that, even in systemic disease, a strategy to remove the primary tumor adds to the improved survival of that patient. Another major advantage of early removal of the primary tumor is the accurate locoregional tumor staging, especially the identification of involved lymph nodes and presence of lymphovascular invasion or T4 local disease. These features are crucial for patient prognosis and thus management for two reasons. First, residual involved lymph nodes after surgery can precipitate local recurrence (especially in rectal cancer). Second, decision making on the need for adjuvant chemotherapy depends on the pathologic stage of the cancer. Chemotherapy offers proven survival benefit (level 1 evidence) for those with stage III tumors. There is some evidence for benefits also in high-risk stage II tumors with features of lymphovascular invasion, emergency presentations with obstruction or perforation, poorly differentiated and T4N0 tumors. Three variables contribute to the yield of lymph nodes: the aggressiveness of surgery, the diligence of the pathologist in searching the specimen, and the anatomy of the patient and the tumor. Evidence shows that when more nodes are pathologically examined (at least 12), tumors are significantly more likely to be classified as node-positive. The implication for patients with limited colorectal liver metastases is that this may influence the timing and sequence of when to integrate systemic chemotherapy. A heavily node-positive pathology (in relation to the primary site) implies that the attempted adjuvant eradication of extrahepatic micrometastatic disease weighs just as heavily in the decision-making process as eradication of the liver metastases, and both concepts need to be integrated in the MDT planning process.

Over the last 35 years, 5-fluorouracil (5-FU)–based adjuvant therapy has become the mainstay of treatment for patients with Dukes C or stage III colon cancer. The standard treatment has been a course of 5-fluorouracil and folinic acid (FUFA), given intravenously (IV) over 6 months. Pooled data from seven randomized controlled trials in patients with stage II or III colon cancer suggest that 5-FU/FA regimens increase disease-free survival at 5 years from 55% to 67% and overall survival from 64% to 71% when compared to surgical resection alone. The potential benefits and risks of chemotherapy should be discussed with patients for whom it is judged appropriate, so that they can make an informed choice. Adjuvant chemotherapy should be scheduled to begin within 6–8 weeks postsurgery. The MOSAIC study was the first trial to show a statistically significant disease-free survival benefit for a treatment regimen other than 5-FU for stage III colon cancer in the adjuvant setting. At 4 years, there was a 25% reduction in the risk of disease recurrence in these patients for the combination oxaliplatin/5-FU/FA compared with 5-FU/FA alone ( P = 0.002). The place of chemotherapy in the treatment of patients with Dukes stage B cancer must be a matter for discussion between patients and their oncologists, as the benefits in some groups are small. The QUASAR1 trial and MOSAIC both showed statistically significant risk reductions in those treated compared to the control arms of no-treatment (best supportive care) and the addition of oxaliplatin to 5-FU, respectively. The survival benefit for patients with rectal cancer is believed to be similar, although the evidence is much weaker, and often confounded by the need to give chemoradiotherapy.