Hepatic Colorectal Metastasis: Current Status of Surgical Therapy
Hepatic Colorectal Metastasis: Current Status of Surgical Therapy
Hepatic colorectal metastases occur in over
50,000 patients in the United States each year, making the liver the
most common site of metastasis from this cancer. Untreated,
hepatic colorectal metastasis is a rapidly fatal disease, with
patients usually surviving for no more than a few months, and
survival beyond 5 years a rarity.[2-6]
Systemic chemotherapy may modestly prolong survival but, even with
the most active agents, rarely extends the patients life for
more than 3 years.[7-15] Surgical resection remains the only
potentially curative option for hepatic colorectal metastases,[16,17]
and as the safety of resections has improved over the last 2
decades,[18-23] liver resection has come to be accepted as the
treatment of choice.
This review summarizes the data demonstrating the safety and efficacy
of liver resection in this clinical setting. The strategy for
preoperative evaluation and patient selection, as well as for
postoperative follow-up, will be presented. Other surgical modalities
utilized in the treatment of patients with colorectal metastases will
be mentioned. In addition, areas of active investigation, such as the
role of adjuvant chemotherapy and comparison of ablation with
resection, will be discussed. (For a discussions of the technical
aspects of liver resection, readers are referred to a recent review.)
Untreated colorectal metastasis to the liver has a poor prognosis and
is associated with a median survival of 5 to 10 months.[2,5,25-28]
Two studies that have specifically analyzed data from patients with
untreated but potentially resectable metastases have reported a
survival rate of 77% at 1 year, 14% to 23% at 3 years, and 2% to 8%
at 5 years.[4,16] Metastatic colorectal cancer to the liver is
therefore a rapidly fatal disease. Although patients with solitary
lesions or unilobar disease appear to have a better prognosis than
those with diffuse disease, 5-year survival is rare among patients
with any untreated disease.
The most successful chemotherapeutic regimens for metastatic
colorectal cancer have been based on fluorouracil (5-FU).[7-11,13-15]
Tumor response rates to 5-FU, 5-FU/methyl-CCNU (semustine), or
5-FU/leucovorin range from 25% to 30%.[7-10] Complete response is
rare, and median survival is generally a year or less. For
unresectable metastatic disease, the current standard therapy is 5-FU
and leucovorin. Chemotherapy does not offer the potential for cure
and therefore is a poor substitute for potentially curative resection.
Over the last 3 decades, an increasingly aggressive surgical approach
to the treatment of hepatic colorectal metastases has been
undertaken. The acceptance of surgical resection as standard
treatment for hepatic colorectal metastases is based on both the
increasing safety of major liver resections and the growing body of
data demonstrating that such resections can be curative.
Perioperative Mortality and Morbidity
Data accumulated over the last 3 decades indicate that aggressive
surgical resections of liver metastases are safe and effective.
Table 1 summarizes data from the series published to date with
more than 100 patients each. From these data, it is clear that
mortality is low even for the most extensive resections; in all
recent series, elective liver resection for colorectal metastases has
been accomplished with < 5% mortality.[18-21,29-36] Perioperative
death usually results from hemorrhage or liver failure.
A relatively high complication rate is still to be expected after
liver resection. This is not surprising given the physiologic stress
of losing a significant portion of an organ as vital the liver. Most
series report complication rates in the range of 20% to 50% (Table
Liver failure is a complication of 1% to 5% of all major liver
resections.[19,20,35,37] Bile leak and biliary fistula occur in
approximately 3% to 4% of all patients,[20,35] and peri-hepatic
abscess develops in 1% to 9%.[18-20,30,35] Hemorrhage is a feared
complication and is a major cause of perioperative mortality but
occurs rarely (1% to 3%). Cardiopulmonary complications include
myocardial infarction (1% in most series),[18-20,30] sympathetic
pleural effusions that may require tube thoracostomy (5% to
10%),[18,38] pneumonia (5% to 22%),[19,35] and pulmonary embolism (1%).[20,37]
As stated previously, the high complication rate of liver resection
for metastatic colorectal cancer does not translate into a high
mortality (< 5% at major centers), nor does it mean that patients
require prolonged hospital stays or frequent admissions to the
intensive care unit (ICU). Median hospital stay even for the most
extensive resection in centers experienced in liver surgery is less
than 2 weeks. In a series of 577 consecutive liver resections
performed at Memorial Sloan-Kettering, including 338 resections
involving a lobe or more, the median hospital stay was 13 days and
ICU admission was required for only 7% of patients.
Surgical resection can also extend survival among patients with
hepatic colorectal metastases. Of patients who undergo liver
resection for colorectal metastases, 25% to 37% can be expected to
survive for 5 years (Table 1), with
median survival between 28 to 40 months.[19,22,23,36,39-42] In a
series with very long-term follow-up, Scheele reported a 10-year
survival rate of 20% after hepatic resection for colorectal metastases.
Thus, there is no longer any doubt that liver resection can provide
long-term survival and cure for patients with metastatic colorectal
cancer. Compared to the results of no treatment (median survival, 6
to 12 months; rare 5-year survival) or of any chemotherapeutic
regimen (median survival, 12 to 18 months; rare 5-year survival),
surgery offers superior results and the only potential for cure.
Consequently, surgical resection has become the standard treatment
for resectable colorectal metastases.
Over the last 2 decades, patient selection criteria based on general
medical condition, extent of systemic disease, and distribution of
liver tumors have been refined to allow for safe and effective resections.
General Medical Condition
As with all major surgery, the general medical condition of the
patient has a major influence on the outcome of liver resection.
Patients with pulmonary compromise are at particular risk for
postoperative complications. This is due to the fact that major liver
resections are generally performed through a long, high, transverse
incision, and postoperative respiratory effort may be associated with
significant discomfort. A sympathetic right pleural effusion also
develops routinely and further contributes to respiratory compromise.
Significant fluctuation of intravascular fluid volume may also occur
perioperatively, putting patients with cardiac disease at risk for
arrhythmias or ischemia. Patients with a history of cardiac,
pulmonary, or other systemic diseases should be assessed carefully
and their medical conditions optimized preoperatively.
In the past, it had been suggested that advanced chronologic age
alone may be a contraindication to surgery. Our most recent
experience does not substantiate this. In a recent report, 128
patients over age 70 years who underwent liver resection had
identical morbidity and mortality as patients younger than age 70 who
were operated on in the same institution during the same time
period. Thus, we do not consider advanced chronologic age alone
to be a contraindication to liver surgery. Nevertheless, we routinely
refer patients over age 65 for cardiopulmonary evaluation prior to surgery.
Extent of Distant Disease
The chest, other sites within the abdomen, and the colon are the most
common sites for additional metastases. For evaluation of
extrahepatic disease, a chest CT is mandatory, as is an abdominal/pelvic
CT. Colonoscopy within 6 months of the exploration for liver
metastases completes the work-up. Bone scans are rarely helpful and
are not done routinely.
Any metachronous colorectal lesion or anastomotic recurrences may be
resected at the time of liver resection. Findings at any other site
of intra-abdominal, extrahepatic metastases usually rule out a liver
resection, as the prognosis is poor and cure is unlikely regardless
of treatment. Disseminated pulmonary metastases also contraindicate
liver resection, although aggressive resection of limited pulmonary
metastases (three lesions) along with the liver metastases has met
with reasonable success.[44,45]
All other imaging modalities to assess the extent of extrahepatic
disease must be considered experimental. Over the last decade,
clinical studies have examined radioimmune imaging using radiolabeled
monoclonal antibodies directed against specific colorectal tumors. At
present, there is no established role for such radiolabeled
antibodies[46-49] in the presurgical evaluation of patients with
liver metastases. Even at centers that advocate the use of such
scans, the false-positive rate is > 10%. These scans are
therefore not sufficiently accurate to preclude surgical therapy.
In contrast, whole-body positron emission tomography (PET) after
administration of [18F]5-fluorodeoxyglucose (5-FDG) shows promise as
a useful test in the evaluation of patients prior to resection of
colorectal metastases.[50-52] Fluorodeoxyglucose is a glucose analog
that accumulates preferentially in certain tumors. In a small study
of 15 patients thought to have resectable hepatic colorectal
metastases, 4 were found by PET to have more extensive disease in the
liver than originally suspected, and unnecessary surgery was
avoided. These results need to be substantiated by larger
Extent of Liver Disease
Metastatic liver tumors are largely silent until the disease is well
advanced. Recent advances in imaging studies must be credited for the
tremendous improvements in outcome of surgical treatments for
metastatic colorectal cancer. These imaging studies permit the
accurate detection of metastatic disease at an early, resectable
stage, and also allow for precise planning of the surgical procedure.
In assessing liver lesions for feasibility of resection, the goals of
imaging are to: (1) characterize the lesions, (2) determine their
distribution, and (3) ascertain their proximity to major vascular and
Transcutaneous and Intraoperative Ultrasound: Transcutaneous
ultrasound is the least invasive, least expensive diagnostic modality
used in the evaluation of liver metastases. This modality is also the
most dependent on the expertise and diligence of the operator. In
addition, overlying air-filled structures, such as bowel and lung
tissue, may obscure the lesions on imaging. For this reason, the dome
of the liver is particularly difficult to assess. Nevertheless, in
expert hands, transcutaneous ultrasound may be as accurate as
computed tomography (CT) or magnetic resonance imaging (MRI) in
determining the number and size of lesions, as well as their
relationship to major blood vessels.
Intraoperative ultrasound is equally operator-dependent but has
become an indispensable tool in the surgical treatment of liver
tumors. This modality is useful both for increasing the detection of
small and/or deep hepatic lesions that are nonpalpable and for
helping to locate intrahepatic vasculature to guide resection.
Computed tomography is the most widely employed imaging test
for the evaluation of liver metastases. Dynamic CT, a technique that
entails the rapid intravenous infusion of contrast agents, is
particularly useful for detecting hypovascular tumors that are
characteristically colorectal carcinomas.
Computed tomographic portography is a refinement of dynamic CT in
which the contrast agent is injected into the superior mesenteric
artery. The contrast material then reaches the portal circulation and
provides particularly good enhancement of the normal liver
parenchyma. Since colorectal metastases are nourished mainly by the
hepatic artery and derive little blood supply from the portal vein,
they appear as filling defects surrounded by hypervascular liver parenchyma.
Computed tomographic portography remains the most sensitive test for
the detection of metastatic colorectal cancer and is the gold
standard for evaluating the number of lesions. The disadvantages of
this test include its invasive nature and cost. Computed tomographic
portography also exaggerates the size of lesions, making it difficult
to assess their proximity to major vasculature.
Magnetic resonance imaging is superb for the identification
and characterization of liver lesions. Metastatic colorectal cancers
are characteristically low-intensity lesions on T1-weighted spin echo
images, and intermediate in intensity on T2-weighted images. These
characteristics allow MRI to distinguish metastatic tumors from
benign liver lesions, including benign cysts, hemangiomas, and
fibronodular hyperplasia. Furthermore, MRI is very useful for the
visualization of vascular structures, such as hepatic veins and vena cava.
With recent advances in magnets and computer software, accurate
imaging of the biliary tree is also now possible with magnetic
resonance cholangiopancreatography (MRCP). When vascular or biliary
involvement is suspected, MRCP provides a single noninvasive test to
assist in surgical planning.
For a patient being considered for resection of metastatic colorectal
cancer, a chest, abdominal, and pelvic CT should be performed.
Computed tomographic portography should be the technique used for the
abdominal CT to allow for the most sensitive detection of liver
tumors. When tumor proximity to major vascular or biliary structures
is apparent, either ultrasound or MRI should be performed, depending
on the expertise available at the local institution.