The liver is the most common site for colorectal cancer (CRC) metastases; 15% to 20% of patients present with synchronous lesions, and metachronous metastases occur in 40% to 50% of patients. Liver metastases are present in 60% to 70% of all patients who die of colorectal cancer. However, management of colorectal liver metastases (CRLM) has changed a lot during the past decade, and overall survival after liver surgery for colorectal metastatic disease has more than doubled. Resection is currently the gold standard of treatment, with 5-year overall survival rates ranging from 27% to 52%, but only 15% to 20% of liver metastases are considered resectable at the time of diagnosis. Patients with untreated CLRM have a median survival of 5 to 12 months.
Diagnosis and Preoperative Workup
Workup begins with a focused history and physical examination to assess the patient’s general functional status, medical comorbidities, and symptoms from the primary colorectal lesion such as obstruction, perforation, or bleeding. Patients with symptoms from the primary tumor typically undergo treatment. Staging studies should be performed before surgery. Laboratory tests are performed, including a baseline carcinoembryonic antigen (CEA) level.
Preoperative staging commonly entails use of a chest computed tomography (CT) scan with use of contrast material and abdominal imaging. Our preference for abdominal imaging is contrast-enhanced liver magnetic resonance imaging (MRI), although triple-phase CT can be equally informative ( Table 63-1 ). For rectal cancer, pelvic MRI is useful to define the extent of the tumor. Positron emission tomographic imaging has not been shown to be an effective screening tool, but it may be useful in determining the nature of ambiguous nodes and liver lesions.
|CT||Triple-phase CT is highly accurate; it offers subcentimeter slices, high-resolution images, and options for vascular reconstruction and volumetrics|
|MRI||Better at identifying and characterizing liver lesions; least affected by postchemotherapy hepatic steatosis; preferred imaging modality in neoadjuvant setting; better at delineating “missing metastasis” after neoadjuvant chemotherapy|
|PET||Significantly superior in detecting extrahepatic spread; PET/CT provides precise localization of a CT and functional data of PET; high costs and restricted availability make it a poor screening test|
|IOUS||Helpful in accurately delineating liver lesions during liver resection or ablation; lesions may be hypo-/hyper-/isoechoic on ultrasound; contrast-enhanced ultrasound, though not yet approved by the FDA, has been shown to detect 97% of lesions seen on helical CT|
Serologic and Molecular Markers
CEA levels are used as a preoperative prognostic indicator in patients with known CRC, as well as postoperatively to detect asymptomatic recurrences of colorectal cancer and to monitor the response of metastases to treatment.
KRAS mutation analysis, with reflex BRAF testing in cases of wild-type KRAS, is used to predict a lack of response to cetuximab and panitumumab in the treatment of metastatic CRC.
A needle biopsy is indicated when diagnosis of a liver lesion is in doubt or in patients with unresectable disease when histologic diagnosis would modify treatment. In the setting of prior CRC, characteristic CT or MRI findings, and raised CEA levels, a needle biopsy may not be necessary and can adversely affect outcome. In a retrospective study, patients who underwent a needle biopsy prior to resection were found to have a higher incidence of needle-track implants and were more likely to have a worse prognosis.
Case discussion involving oncologists, pathologists, radiologists, colorectal surgeons, and a surgeon experienced in the treatment of hepatobiliary and extrahepatic metastases is an essential part of treatment planning.
Staging and Prognosis
The following staging system for metastatic disease was proposed by European Colorectal Metastases Treatment Group in 2007:
M0: No metastases
M1a: Resectable liver metastases
M1b: Potentially resectable liver metastases
M1c: Liver metastases that are unlikely to be resectable
Poor prognostic factors for liver resection include more than four liver tumors, poorly differentiated tumor, bilobar lesions, bulky lesions, hilar/celiac lymph nodes, inoperability of the primary tumor, local recurrence, lung metastases, other extrahepatic metastases, age older than 70 years, and cardiovascular/respiratory disease. The presence of portal nodal metastasis has been consistently associated with poorer prognosis, particularly when the hepatic artery group of nodes is involved. These patients appear to benefit most from multimodality treatment, including chemotherapy followed by surgical resection. When tumors are discovered intraoperatively, tumor response to chemotherapy has been used to guide further resection.
Postoperatively, infectious complications are associated with decreased overall survival. Bilobar disease, more than eight tumors, more extensive treatment than a hemihepatectomy, and R1 margin status have been associated with increased postoperative complications.
Over the years, the following scoring systems have been proposed to help with risk stratification and patient selection for liver resection:
Fong et al—scoring system (1999)
Iwatsuki et al (Pittsburgh risk score, 1999)
Nagashima et al—scoring system (2006)
Zakaria et al (2007)
Rees et al (2008)
These scoring systems were analyzed by Gregoire et al in 2010, who concluded that no single prognostic factor was common for all the scores. However, the most common factors included the number and size of the largest metastases, the disease-free interval between diagnosis of the primary cancer and development of liver metastases, CEA level, a stage III primary cancer, and presence of extrahepatic disease. It has become clear that generalized criteria may not be appropriate for every patient and that a more individualized approach is warranted.
Initial treatment of a patient presenting with CRLM depends on his or her presentation. A useful starting point is to determine if the patient has symptoms relating to the primary tumor. In patients with signs and symptoms suggestive of hemorrhage, perforation, or obstruction, initial treatment should be resection or palliation of the primary tumor. In an asymptomatic patient, the initial treatment should be discussed by a multidisciplinary tumor board.
Chemotherapy for CRC started with fluoropyrimidines in the 1950s, which significantly improved median overall survival rates. The combination of 5-fluorouracil (5-FU) and leucovorin was a significant advance, and combination regimens of infusional 5-FU with either oxaliplatin (FOLFOX) or irinotecan (FOLFIRI) currently are the mainstay of CRLM management. Targeted therapy using biologic agents is mostly used as an adjunct to systemic chemotherapy. Bevacizumab (anti–vascular endothelial growth factor) has been shown to improve overall survival and protects against oxaliplatin-induced injury, and cetuximab (anti–epidermal growth factor receptor) is effective against tumors containing wild-type KRAS . Panitumumab (anti–epidermal growth factor receptor) and Perifosine (anti-AKT and nuclear factor–κB) are other examples of biologic therapy.
Neoadjuvant Chemotherapy for Resectable Liver Disease
Preoperative chemotherapy for resectable liver metastasis has been a matter of debate. The principles of neoadjuvant chemotherapy are to eradicate micrometastases ( Table 63-2 ) and shrink tumors to a more resectable size. However, the response of tumors to preoperative chemotherapy correlates with improved overall survival. The EORTC 40983 prospective randomized clinical trial showed an improvement in progression-free survival in patients who underwent perioperative chemotherapy (oxaliplatin) combined with resection compared with the group having surgery alone (36.2% vs 28.1%, P = .041), and National Comprehensive Cancer Network guidelines now recommend 6 months of perioperative chemotherapy for patients with resectable CRLM.
|Type of Chemotherapy||Advantages|
|Decreases tumor load; vascular pedicles may become free of tumor; large lesions may shrink enough for radiofrequency ablation (<3 cm)|
|Allows limited hepatectomy, treatment of micrometastases, evaluation of possible response to further chemotherapy|
|Treatment of remnant “dormant” tumor cells after resection; decreases recurrence rates|
|Localized, targeted therapy using drug-eluting beads; useful adjunctive therapy in persons who do not respond to first- and second-line chemotherapy|