Abeloff's Clinical Oncology, 4th Edition

Part II – Problems Common to Cancer and its Therapy

Section F – Local Effects of Cancer and Its Metastasis

Chapter 59 – Liver Metastases

Nancy Kemeny,Margaret Kemeny,
Laura Dawson

SUMMARY OF KEY POINTS

Incidence

  

   

Liver metastases are most frequently seen in patients with colorectal cancer (CRC; nearly 15% of patients presenting and an additional 60% developing subsequent spread); they are less common in patients with breast cancer (4% of initial failures), lung cancer (15%), and melanoma (24%).

Etiology

  

   

The liver has a rich blood supply from both the hepatic artery and the portal vein; metastases can reach the liver from any organ, but the direct passage of blood from the gastrointestinal tract to the liver via the portal circulation plays a critical role in explaining the high rate of liver metastases from these sites.

Detection

  

   

Contrast computed tomography (CT) and magnetic resonance imaging (MRI) can detect approximately two-thirds of liver metastases.

  

   

CT angiography, CT portography, and intraoperative ultrasound seem to have increased sensitivity as compared with standard techniques.

  

   

Positron emission tomography (PET) is more useful to detect extrahepatic disease.

  

   

Of laboratory tests, carcinoembryonic antigen (CEA) can be useful for patients with metastatic CRC to the liver.

Treatment

Hepatic Resection

  

   

There is general agreement that surgical resection is the treatment of choice for patients with one to three metastases from CRC, producing a 5-year survival of about 30%.

  

   

Significant advances in surgical technique include the ability to perform metastasectomies (rather than formal lobectomies) and total vascular exclusion.

  

   

The liver is the chief site of relapse after hepatic resection (50% of all patients).

  

   

Prognostic variables that influence survival after hepatic resection include the presence of extrahepatic disease, the stage of the primary colon cancer, the time interval between primary and the development of hepatic metastases, and the number of metastases and positive margins.

  

   

The use of adjuvant hepatic artery infusion (HAI) after liver resection has produced positive results in two American trials. There is a clear decrease in hepatic recurrence with adjuvant HAI.

Systemic Chemotherapy

  

   

The objective response rate with intravenous chemotherapy is improving, with response rates of 40% to 50% among patients with breast cancer, gastric cancer, and now even colon cancer.

Hepatic Artery Infusion

  

   

Eight randomized trials demonstrate a higher response rate for HAI than for systemic infusion (42% to 64% vs. 0 to 38%, respectively). Survival advantage is difficult to interpret: two large U.S. studies allowed a crossover from systemic therapy to HAI after tumor failure. New CALGB study without crossover showed a survival advantage. The chief toxicity of HAI is biliary enzyme elevations in 40% and biliary sclerosis in 5% to 35% of patients.

Hepatic Artery Embolization

  

   

Embolization could play a role in highly vascular tumors (neuroendocrine tumors and hepatocellular carcinoma, HCC).

  

   

Embolization agents include Gelfoam, lipiodol, and degradable starch microspheres.

  

   

Embolization is rarely useful for patients with metastatic CRC.

  

   

Chemoembolization involves the local entrapment of drug in the embolization agent in an attempt to provide a prolonged exposure of the tumor to drug locally, with less systemic exposure.

Ablative Techniques

  

   

Cryosurgery involves destruction of tissue using a freezing probe.

  

   

Limitations include the difficulty of controlling freezing and the typical requirement for laparotomy.

  

   

Radiofrequency ablation involves destruction by frictional heat and can be used percutaneously but is rarely good for lesions larger than 3 cm.

Absolute Ethanol Injection

  

   

Ethanol is injected into the tumor under ultrasound guidance and could be of use for small HCCs.

Radiation

  

   

Whole-liver external-beam irradiation therapy alone is limited by the occurrence of radiation hepatitis to about 30 Gy in 15 fractions.

  

   

Parts of the liver can be treated to far higher doses using yttrium-90 (90Y) microspheres, interstitial brachytherapy, and external-beam irradiation therapy guided by three-dimensional treatment planning.

INTRODUCTION

Not so long ago an oncology textbook would not have devoted an entire chapter to liver metastases. Oncologists were so pessimistic about the appearance of such metastases that “no treatment” was often the recommendation. Enormous changes have occurred, so that now, diagnosing liver metastases early can lead to effective treatment and even cure for a growing percentage of patients.

The liver is the primary site of metastases for many malignant neoplasms. Gastrointestinal malignancies are especially prone to spread to the liver because of its portal venous drainage. Extra-abdominal tumors such as bronchogenic carcinoma, breast cancer, and malignant melanoma often spread hematogenously to the liver.

For gastrointestinal tumors, differences are seen in the natural history of the hepatic metastases. In some circumstances, hepatic metastases are a sign of disseminated disease. When gastric and pancreatic cancers metastasize to the liver, the mean survival is short, and widespread metastases often exist, so that radical measures such as hepatic resection or hepatic artery infusion (HAI) are rarely appropriate. In contrast, for colorectal cancer (CRC) the liver might be the sole site of metastatic disease, and a sizeable number of these patients might have isolated liver metastasis. In this setting, progress has been significant in the areas of hepatic resection, regional chemotherapy, and radiation therapy as discussed in this chapter.

For nongastrointestinal tumors, metastases to the liver are less common as the initial site of relapse. Although breast, lung, and melanoma are the main extragastrointestinal cancers to metastasize to the liver, initial isolated metastases in the liver occur in 4%, 15%, and 24% of these patients, respectively.

Treatment of these types of metastases varies according to the sensitivity of the tumor type to chemotherapy. Those that are more sensitive to antineoplastic agents might benefit from systemic therapy or aggressive regional approaches (breast), whereas those that have limited response to chemotherapy (such as melanoma) may be approached regionally in the setting of a clinical trial. Special care must be exercised in choosing patients for regional therapy who do not have CRC, because diseases such as breast cancer and melanoma are rarely confined to the liver.

The last 20 years have witnessed new and more accurate methods of detecting and quantifying liver metastases, more advanced surgical techniques facilitating hepatic resection, and biologic advances that have increased the spectrum of available regional therapies. This chapter discusses advances in detection and treatment of liver metastases.

DETECTION

The need for early and accurate detection of liver metastases has become more critical as the guidelines for resectability of liver metastases have become more liberal. Furthermore, ablative therapies and conformal radiation therapy are more likely to control unresectable liver metastases detected at an earlier stage, with a lower burden of disease. The most common imaging modalities for liver metastases imaging and their usefulness are described in the sections that follow.

Imaging Techniques

Computed Tomography

Computed tomography (CT) has been used for more than 20 years to image hepatic metastases. Modern CT scanners are capable of rapid scanning and high-resolution three-dimensional images of the liver that can be acquired during one breath hold.[1] Projection and reformatted views can be useful when assessing the proximity of liver metastases to the vessels. CT images are routinely evaluated by two window levels to maximize detection of lesions. A soft-tissue window (width of 300–500 Hounsfield units) is used for the initial examination of the liver and adjacent abdominal architecture. A second setting with a narrow window and a lower width (100–150 Hounsfield units) is then used to evaluate the liver, because this setting increases contrast differences between the normal liver parenchyma and abnormalities.

Noncontrast Computed Tomography

Contrast CT is sometimes not possible because of contrast allergic reactions or renal impairment. Although the sensitivity and specificity of noncontrast CT is far reduced as compared to contrast CT, it may help in identifying hypervascular metastases (especially carcinoid tumors, islet cell tumors, and renal cell carcinomas) or visualizing calcifications or hemorrhage. Noncontrast CT often fails to distinguish hypovascular tumors from the liver parenchyma. Nonenhanced blood vessels may also appear as low-attenuation masses and be confused with metastases.[2]

Contrast Computed Tomography

Intravenous contrast is infused over approximately 2 min, with repeat imaging acquired during the infusion to capture the different enhancing phases of tumors. The two most useful phases for hepatic metastases are the arterial phase (25-sec delay) and portal venous phase (60-sec delay). During portal venous imaging, the liver parenchyma enhances and hypovascular metastases, such as those from CRC, appear as filling defects.[2] Most metastases are seen best in the portal venous phase, but some are best seen in delayed venous and occasionally arterial phases. Hypervascular liver metastases may be missed with CT, because very vascular metastases may enhance to the same degree as normal liver parenchyma. Delayed contrast CT, referring to scanning 4 to 6 hr after contrast injection, is most useful as an adjunct to increase the sensitivity and accuracy of contrast CT, because malignant lesions usually do not retain contrast and appear as hypodense areas within the enhanced normal liver parenchyma.[2]

Computed Tomography Angiography and Portography

The rapidity with which multidetector CT scanners can acquire images makes possible imaging of the entire liver during the peak of the bolus of intravenous contrast, supporting CT angiography and portography.[3] This noninvasive imaging can provide images of hepatic vasculature of superb quality, avoiding catheter angiography. This technique is useful to detect vascular anomalies that may help in planning of hepatic resection or hepatic arterial chemotherapy. Most often, imaging is acquired during an arterial phase and a portal venous phase.

Magnetic Resonance Imaging

Magnetic resonance imaging (MRI) exploits differences in magnetic properties of atomic nuclei to produce images. When placed in a magnet, protons have a nuclear spin that aligns with the magnetic field and give off radiowaves (MR signal) on returning to equilibrium. As the chemical environment changes, there is a change in the frequency of the signal, and the image. Differences in the time to return to equilibrium (relaxation times) are exploited in MRI, using T1- and T2-weighted pulse sequences. Most commonly, axial imaging is used for liver metastases imaging; however, coronal or sagittal views can define better a tumor's proximity to adjacent vessels. The most common pulse sequences used for liver metastases imaging are a T1- or T2-weighted spin echo. The T1 images generally show metastases as low-intensity lesions, whereas in T2-weighted images, metastases are of high signal intensity ( Fig. 59-1 ). T2-weighted sequences are generally superior for detection and characterization of liver masses. Benign cysts and hemangiomas usually appear homogeneous and bright, whereas metastatic lesions are less bright and more heterogeneous. [2] [4] Although MRI inherently provides for soft-tissue contrast, contrast agents, such as gadolinium (Gd) chelates, can improve diagnostic accuracy. Similar to CT, MR angiography can be used as a noninvasive method to evaluate hepatic vasculature. Novel MR contrast agents have the potential for improving detection of liver metastases.[5] Advantages of MRI include the lack of radiation exposure to the patient and the low frequency of reactions to contrast agents. MRI is particularly useful for patients with contraindications to iodinated contrast, and MRI can detect hypervascular lesions that are not well visualized on CT. Limitations of MRI include inherent warping and the time required for imaging, which may lead to artifacts in patients who cannot hold their breath. Parallel imaging allows for high-resolution imaging of the entire liver within one breath hold, with the potential for more sophisticated sequences (such as diffusion-weighted imaging and MR spectroscopy) to improve liver metastases detection in the future.

 
 

Figure 59-1  A, T2-weighted spin-echo image (left); T1-weighted spin echo image (right). In T2-weighted images the metastases are areas of high signal intensity. In T1-wieghted images metastases have a low intensity. B, Dynamic gadolinium-enhanced MRI demonstrating peripheral nodular enhancement in a hepatic hemangioma. Enhancement increases as the study advances over time from upper left to lower left and then to upper and lower right.

 

 

Ultrasonography

Ultrasonography (US) relies on sound waves to generate an image. US is most commonly used for screening for metastases because of its wide availability and lack of the need for radiation exposure. It is most useful in detecting superficial liver metastases in small patients. Contrast-enhanced US, using intravascular microbubble contrast agents, has shown similar accuracy for liver metastases detection compared to CT and MR.[6] An advantage of contrast-enhanced US is the potential for characterization of liver lesions based on morphologic evaluation as well as temporal vascular enhancement pattern.[7] During the portal venous phase, benign lesions typically enhance more than the liver, whereas malignant lesions enhance less.[8] Expertise in contrast-enhanced US is not yet widespread.

Intraoperative US can also be useful at detecting small, deep hepatic metastases not palpable. In a study of 84 patients undergoing colon resection, intraoperative US detected 14 hepatic metastases that were missed on surgical palpation.[9] In a similar study, intraoperative US detected seven nonpalpable lesions in 70 patients.[10]

Positron Emission Tomography

Positron emission tomography (PET), in which a radioactively labeled tracer is administered to the patient and the scanner collects the emitted positron radioactivity to generate an image, allows imaging of cellular processes (such as cellular proliferation (18F-labeled thymidine), hypoxia (18F-labeled Miso), and blood flow ([15O]water) to be visualized. The majority of clinical experience has been with fluorodeoxyglucose ([18F]FDG), which images cellular glucose metabolism. A limitation of PET scans of the liver is that they are acquired over many breathing cycles, leading to blurring due to respiratory motion and lower spatial resolution as compared with CT. This has driven the development of dual-modality scanners capable of CT and PET scanning. Scanners capable of respiratory sorting are also being developed.

FDG PET scans have been done to stage patients before resection of liver metastases [11] [12] [13] and to evaluate response to treatment.[14] In one study, 6 of 34 liver metastases treated with chemoembolization demonstrated persistent FDG PET activity, which led to further treatment.[15] FDG PET has also been found to be useful in diagnosing peritoneal recurrences in patients with elevated serum carcinoembryonic antigen (CEA), with a sensitivity and accuracy for peritoneal disease detection of 88% and 78%, respectively (as compared with 38% and 44% for CT).[16] In patients with colorectal liver metastases staged planned for resection, Ruers and colleagues found that 10 of 51 patients (20%) had a change in treatment plan due to unresectable metastases or extrahepatic metastases detected on FDG PET.[12] Other studies have also demonstrated a change in treatment resulting from FDG PET scans obtained before planned liver resection in 20% to 58% of patients. [13] [17] [18] [19] In contrast, 15 out of 52 pathologically proven liver metastases were missed with FDG PET in another study, emphasizing that FDG PET cannot be the sole method used to evaluate liver metastases. Five-year survival of a series of 100 patients with liver metastases, imaged with FDG PET and treated with hepatic resection, was 58% (46% to 72%), demonstrating excellent outcomes in well-selected patients.[20]A continuing Canadian randomized trial of PET versus no additional imaging in resectable liver metastases, funded by the Ontario Clinical Oncology Group (OCOG), has the potential to more clearly define which patients with liver metastases planned for resection are most likely to have a change in treatment and benefit from PET scans obtained before resection (personal communication, Steven Gallinger, Toronto, March 2007).

Comparison of Modalities

Many studies have compared imaging modalities for the detection of liver metastases. A German study[21] prospectively compared dynamic CT with MRI and US in 75 patients with known gastrointestinal tumors before exploratory laparotomy. Ninety-five liver metastases were detected in 32 patients.[22] Of the 95 lesions, 68% were detected by CT, 63% by MRI, and 53% by US. Another prospective study involving 69 patients came to the same conclusion, and also demonstrated the superiority of MRI T2-weighted images over T1.[23] Other studies have shown improved detection of liver tumors with MRI,[24] [25] and differences in the shape and size of hepatic metastases depending on which CT or MR sequence is used for imaging.[26] Overall, CT scans and MRI seem to have similar sensitivity and specificity for detection of liver metastases using state-of-the-art techniques at experienced centers.

Noncontrast external US has a lower rate of accuracy for liver metastases detection as compared with contrast CT or MRI. A unique study of US before liver transplantation allowed the accuracy to be determined pathologically.[27] Only 36 of 80 tumors in 34 patients were detected, indicating a sensitivity of 45%. The majority of lesions not detected were smaller than 1 cm. Contrast-enhanced US has demonstrated similar accuracy for liver metastases detection compared to CT and MRI.[6]

FDG PET can detect disease not seen on CT or MRI in patients with liver metastases. [12] [13] [28] In a comprehensive review of the literature,[11] the sensitivity and specificity of FDG PET for hepatic disease detection were 88% and 96% and for extrahepatic disease detection were 92% and 85%, respectively, as compared to 83% and 84% for hepatic disease and 61% and 91% for extrahepatic disease in CT. The management was changed based on FDG PET scans in 32% of cases on average. Overall, most studies have confirmed increased sensitivity of FDG PET as compared with MRI and CT, with increased sensitivity on a per-patient analysis as compared with a per-lesion analysis.[5] FDG PET is less sensitive for detection of small liver metastases (less than 1–2 cm) than state-of-the-art MRI or CT.[5] [29] [30]

At present, the standard of care for imaging of hepatic metastases is helical multiphasic intravenous contrast CT scan, with MRI or contrast-enhanced US alternatives for patients with contraindications to CT contrast. For patients whose treatment plan includes resection, extra imaging should be considered to help best select patients with liver-confined metastases.

Differential Diagnosis for Liver Metastases

The majority of liver metastases enhance on the portal venous phase of imaging. Arterial and portal venous phase imaging, using CT, MRI, or US, can help in distinguishing malignant from benign lesions and hepatic metastases from primary cancers. However, several benign lesions cannot always be reliably distinguished from metastases. Hemangiomas may be confused with highly vascular tumors and cysts. Dynamic Gd-enhanced MRI can help distinguish hemangiomas from tumors. Hemangiomas can also be mistaken for metastases in noncontrast CT scans and on portal phase contrast CT (see Fig. 59-1 ). Although portal phase contrast CT is sensitive in detecting small hepatic lesions, it has a high false-positive rate because benign lesions such as hemangiomas, cysts, adenomas, or flow artifacts may be confused with metastases. Technetium-99 (99mTc) red blood cell scintigraphy is the most specific noninvasive test to diagnose hepatic hemangioma ( Fig. 59-2 ).[31] Table 59-1 lists some lesions that may be confused with liver metastases and suggests the tests to differentiate among those entities. A rare entity that can on occasion be confused with metastases is focal fatty infiltration of the liver ( Fig. 59-3 ). This may be seen in patients receiving hyperalimentation. MRI is helpful for distinguishing these lesions.

 
 

Figure 59-2  Hepatic hemangiomas. A, Transaxial single-photon-emission computed tomography (SPECT) image obtained 2 hours after injection of 99mTc-labeled red blood cells. Normal blood activity is seen in the liver, and increased blood pooling activity is seen in the aorta and spleen (which is normal). There is a large area of increased blood pooling activity in the posterior section of the right lobe of the liver that corresponds to an area of reduced attenuation on the CT scan. B, The CT scan is obtained from the same plane as the SPECT scan. An increase in blood pooling activity with an increase in intensity from earlier images is specific for hemangioma.

 

 


Table 59-1   -- Benign Liver Lesions That May Be Confused with Malignancy

Technique

Benign Lesions

Malignancy

MRI (T2-weighted)

Vessels

Small metastatic lesions

 

Hemangiomas

Highly vascular tumor (islet cell, renal, carcinoid)

 

Cysts

Cystic tumors

 

Cysts

Highly vascular tumors (islet cell, renal, carcinoid)

CT

Fatty infiltration

Metastases

 

Fatty infiltration

Hepatocellular cancer

 

Nonenhanced vessels

Metastases

 

Adenoma

Abscess

CTP

Hemangioma

Metastases

 

Cysts

 

 

Adenoma

 

 

Flow artifacts

 

Angiography

Focal nodular hyperplasia

Metastases

 

Hemangioma

 

CT, computed tomography; CTP, CT portography; MRI, magnetic resonance imaging.

 

 

 

 
 

Figure 59-3  Fatty infiltration. This lesion was thought to represent metastatic disease; surgery revealed it was only fatty replacement.

 

 

Biochemical Laboratory Tests

Unlike imaging techniques, the laboratory tests that are available for liver function assessment are not very sensitive. The basic liver function tests include the alkaline phosphatase, bilirubin, albumin, prothrombin time, lactate dehydrogenase (LDH), and serum transaminases. For patients who have metastatic colon cancer, CEA is also extremely useful (see Fig. 59-3 ). Several studies have looked at the usefulness of these tests to detect liver metastases, especially metastases from CRC. [32] [33] [34] CEA remains the most sensitive test for CRC, but even this test can be normal in the presence of liver metastases, especially with minimal hepatic disease. In a prospective study at the City of Hope Hospital (Durante, CA, USA) in patients with metastatic liver disease deemed resectable by CT, the average alkaline phosphatase and LDH levels were within normal limits, whereas CEA was elevated in 73% of the patients.[35] Serum LDH is useful as a prognostic indicator, with high serum LDH denoting a poorer survival.[36]

HEPATIC RESECTION

Because of recent advances in techniques for liver surgery, hepatic resection for metastatic disease has become increasingly safe and used more frequently in treatment over the last two decades. Hepatic resection of metastases was first attempted just before World War II. Experience gained from trauma centers during the war led to the emergence of techniques applicable to resection of metastatic hepatic lesions. In recent years many new devices have helped in the technical aspects of the surgery and in the overall safety of the procedures.

In the last few decades hepatic resections have gone from formal lobectomies to resections along nonanatomic lines. Metastasectomies, or removal of the tumor plus a rim of normal hepatic tissue, can be done rather than a formal lobectomy. With the greater familiarity of the hepatic anatomic segments of Couinaud ( Fig. 59-4 ),[37] current resections are more frequently being performed along these lines especially with the use of intraoperative US. These segment divisions cannot be seen by the naked eye on the surface of the liver but can be mapped out with the help of intraoperative US. In the modern operating room the use of intraoperative US is a necessary skill for the operating surgeon. This technique allows for both a reduction in the amount of normal liver removed and resection of disease from both lobes.

 
 

Figure 59-4  Couinaud's eight hepatic segments.  (From Iwalsuki S, Sheahan DG, Starzl TE: The changing face of hepatic resection. Curr Probl Surg 1989;25:281.)

 

 

 

Several new devices have been introduced to help with the resection itself. Probably the most commonly used device is the cavitron ultrasonic aspirator, which allows dissection of the liver parenchyma without entering the bile ducts. A new technique for dividing the parenchyma is the multiprobe bipolar radiofrequency device, which was introduced in 2004. This device uses radiofrequency waves to coagulate vessels and bile ducts in the liver parenchyma before dividing the tissue.[38] Other devices are available, but in the end it depends on the operating surgeons and their own techniques, which vary from center to center. No clear advantage has been established for one product over the other.

The operative technique of total vascular exclusion for hepatic resections involving tumors near to the vena cava was introduced in the 1980s by surgeons from the liver transplant community. With this technique the vena cava is cross-clamped in two areas (suprahepatic and subhepatic), and the porta hepatis is also occluded. [39] [40] In Bismuth and associates’[41] report on 54 patients using total vascular exclusion, the average duration of cross-clamping was 46 min, and the average operating time was 6 hr. The transfusion requirements were low, and no intraoperative deaths occurred. The application of vascular occlusion seemed to be useful for tumors close to the vena cava or for large central tumors. Again, it is a technique to be used in special circumstances by surgeons who are familiar with the procedure.

The use of the autotransfuser, a technique pioneered in trauma patients, had been avoided initially for patients with malignancies because of fear that tumor cells might be disseminated into the bloodstream with this instrument. Prospective studies on patients undergoing hepatic resection for tumors have not shown this to be true. [42] [43] Animal studies also support the concept that metastatic cells are organ specific and would not disseminate if introduced into the bloodstream. Because of these studies, many centers are using the autotransfusers during hepatic resection and reducing the need for multiple blood transfusions.

Another technical advance for hepatic surgery has been the use of fibrin glue, which can aid in sealing the large, raw surfaces of the liver left after major resections.[44]

The question of drainage after hepatic resection has been addressed prospectively in a study that showed no difference in complications in the drained group versus the undrained group. When the data are analyzed more carefully, however, it becomes clear that patients with lobectomies or greater did benefit from drains, whereas those patients with smaller resections did not need drains placed. [45] [46]However, this study was done before the advent of fibrin glue, which may reduce the need for drains.

In general, a major surgery such as a liver resection has been reserved for a situation in which the operation can be curative. Experience has shown that the resection of colorectal metastases to the liver can be curative in at least one-quarter of patients with certain requirements. The curability by resection of liver metastases from other primary cancers is not quite as clear. The resection of metastases from other gastrointestinal malignancies, such as stomach and pancreas, has been disappointing because of the aggressive nature of these tumors by the time they become metastatic. [47] [48]

The performance of liver resections for metastatic disease from a breast cancer is controversial. Liver metastases from gastrointestinal tumors can be considered regional spread, but for the metastases to go from the breast to the liver requires release of the tumor cells into the systemic circulation. Thus, the concept that the liver could be the only site of spread is harder to prove, and as a result, curative intent is more difficult to achieve. Most of the literature about hepatic resections for breast cancer metastases has come from France. In a recent publication one hospital retrospectively reviewed their experience in 108 patients from the time of 1984 to 2004. Twenty-three (21%) of the patients were found to be unresectable at laparotomy because of intraabdominal metastases or hepatic lesions that were technically not resectable. Of the 85 patients who had a liver resection, 38% had a solitary metastasis, 32% had two or three lesions, and 31% had more than three lesions. The metastases were small with an average size of 2.8 cm. The median followup of the patients was 38 months, with a median survival of 32 months. Only 18 patients were alive after 5 years. A multivariate analysis of factors showed only response to chemotherapy, the absence of extrahepatic metastases, and a clear margin at resection correlated with survival.[49]

Another Parisian team reported their experience from 1988 to 1997 with 49 liver resections for metastatic breast cancer. They reported no mortality and a morbidity of 11.5%. The 1-, 2-, and 3-year survival rates were 86%, 79%, and 65%, respectively. Recurrence in the remaining liver was seen in 49% of the patients at 3 years after resection. Recurrent disease anywhere in the body was seen in 63.8% of patients at 3 years after the resection. The only factor that correlated with survival was the disease-free interval between diagnosis of the primary breast cancer and the appearance of the liver metastases, with a 3-year survival of 45% if the liver metastases appear in less than 4 years after the primary breast cancer versus a 3-year survival of 82% if the breast cancer was more than 4 years from the appearance of the liver metastases.[50]

Another study from Paris reported on a smaller number of resections of solitary hepatic metastases from breast cancer.[51] Of the 32 patients with isolated liver metastases, 27 were found to have actual metastatic disease, whereas 5 had benign disease. Six of the 27 had diffuse disease that was not amenable to resection, whereas of the 21 who underwent hepatic resection, the average survival time was 26 months. For both of these studies there is no way of judging the usefulness of hepatic resection versus chemotherapy for these patients. It is interesting to note that the more recent study has a superiorsurvival time. This could signify several factors, including better drugs and improved selection. That the disease-free survival (DFS) is still quite low, however, emphasizes that even with strict patient selection, the chance of cure with a liver resection is quite limited, but real. Thus patients who have metastatic disease to the liver due to breast cancer need to be screened carefully before they are offered resection as a therapeutic option. Issues such as other sites of metastatic disease, response to chemotherapy, and disease-free interval between breast cancer and liver metastases all must be considered before hepatic resection.

Gastrointestinal neuroendocrine tumors frequently metastasize to the liver but their growth is often slow, and patient survival can be prolonged even without surgical intervention. Both the rarity of these tumors and their prolonged course make them poor candidates for prospective trials. Thus, the role of hepatic resection for metastatic neuroendocrine tumors can only be evaluated in a retrospective manner. At the Mayo Clinic, over a 20-year period from 1970 to 1990, only 37 patients with hepatic metastases from a neuroendocrine tumor had a resection (17 curative resections and 20 palliative resections).[52]Eleven of the 17 patients with curative resections were alive 1 to 92 months after surgery (median 19 months) without evidence of disease. Of the 20 patients who underwent palliative resection, 1 died in the postoperative period, 8 died of disease 9 to 76 months after resection, and 19 patients had some relief of symptoms. The authors concluded that resection is reasonable in cases in which the bulk of tumor can be removed and the patients are symptomatic. A second report from the Mayo Clinic reviewed their hepatic resections for neuroendocrine tumors from 1984 to 1992, with a total of 74 cases.[53]Because this is a report from the same institution and there is some overlap of the time period, clearly many more resections were done in the later years (1990–1992) than in the two decades before. The study included patients whose primary tumors were either completely resected or potentially completely resectable. Patients with carcinoid syndrome were injected with subcutaneous somatostatin preoperatively. Most patients[54] had nonanatomic resections, although 36 had a lobectomy or greater. The mortality rate was 2.7%, and the morbidity was 24.3%. Overall survival at 4 years was 73%, with a mean followup of 2.2 years. All of the 12 patients who died had tumor progression. There was no significant survival difference among those patients who had curative resections (that is, removal of all gross disease) versus those who had palliative resections. Symptom relief was seen in 90% of patients. The authors concluded that resection should precede hepatic arterial occlusion and systemic chemotherapy, because it provides an excellent response rate and good survival. Other studies with considerably fewer patients also support this conclusion. [55] [56] [57]

A review of the literature in 2002 reported on 227 patients with hepatic surgery for metastatic malignancies. The operative mortality for the 212 patients with carcinoid tumors was 2.3% and morbidity was 71%. The overall 5-year survival was 71%, and when looking at only the metastatic islet cell tumors, the survival was even higher at 82%. Symptomatic relief from carcinoid syndrome was 86% with a duration of 4 to 120 months. Some discussion of ablative techniques was offered, but only with very preliminary results. The analysis of these sorts of data suggest that for selected patients liver resection is safe, appropriate, and can result in long-term survival and relief from endocrinopathy symptoms. Guidelines for resection include the preoperative assessment that the primary and metastatic disease is completely resectable ( Box 59-1 ).[59]

Box 59-1 

SURGICAL GUIDELINES FOR HEPATIC RESECTION

The existence of extrahepatic intraabdominal metastases should be excluded before attempting a hepatic resection. Because the periportal lymph nodes are the most common site for intraabdominal extrahepatic metastases, biopsy samples from them should be taken and sent for frozen section. If extrahepatic disease is present the hepatic resection should not be carried out.

The extent of a hepatic resection can span from one small nodule to a trisegmentectomy by which 75% of the liver is removed. The assessment of patients for trisegmentectomy is very difficult, because no tests are currently available to delineate accurately which patients can survive with a 75% loss of liver mass. In general, however, if a patient has cirrhosis, an extensive resection is discouraged. Debilitated (poor performance status) patients are not good candidates for major hepatic resection. Age alone should not preclude a patient's eligibility for hepatic resection.

The other noncolorectal tumors metastatic to the liver that have been resected enough to have reported studies are the sarcomas. The largest retrospective review comes from M.D. Anderson Cancer Center with 66 patients having liver surgery for metastatic sarcoma. The mean size of the lesions was 3.9 cm and mean number of lesions was 3. The majority of patients had a gastrointestinal stromal tumor (GIST) (36 patients), 18 had a leiomyosarcoma, and the rest had various other sarcomas. Thirty-five patients had resection, 13 patients had radiofrequency ablation (RFA), and 18 patients had the combination of resection and RFA. The operative mortality and morbidity were 4.5% and 15.2% respectively. The recurrence rates were high with 85% of patients, with RFA recurring, 89% with the combination of resection and RFA recurring, and 57% of patients with resection only recurring (statistically better for this group). The 5-year DFS and overall survival were 16% and 27%, respectively. The article discussed the issue of the GIST tumors and the targeted therapy with Gleevec (imatinib mesylate). The authors felt that resection was still recommended for the GIST tumors, because the rate of complete response with imatinib mesylate was only around 5%, and the bulk of the patients progressed after 2 years on treatment. The conclusions of the report were that although recurrence is high, hepatic resection in selected patients is indicated and will lead to longer survival.[60]

The group at Memorial Sloan-Kettering (MSKCC) reviewed their experience with hepatic resections for sarcomas in the period from 1982 to 2000. There were 331 patients with liver metastases from a variety of primary sarcomas, 56 of whom had a hepatic resection. Thirty-four of the 56 patients had a GIST. Ten of the 56 patients have actually survived for 5 years, but only 2 are disease free. The disease-specific survival rate for 3 and 5 years was 50% and 30%, respectively. The patients with GIST had the same survival as those with other types of sarcoma. The time interval (less than or greater than 2 years) between the appearance of the primary sarcoma and that of the liver metastases had a significant influence on survival and was the only independent prognostic variable in a multivariate analysis.[61]

An analysis of hepatic resection for leiomyosarcomas from Germany reported a 40-month median survival in patients who had all tumor removed with negative margins. The conclusions from this study, similar to the other reports, were that hepatic resection for leiomyosarcomas could be done safely and in selected cases could prolong survival. Most of the other reports in the literature about resection of hepatic metastases from sarcomas are anecdotal, with far fewer patients in each study. [62] [63]

The use of hepatic resection for metastases from melanoma is moderately rare as seen in a report from two major melanoma centers, where 1750 patients with hepatic metastases from melanoma were identified and only 34 of them had a surgical exploration with the intent to resect the tumor from the liver. Of the 34 patients explored, 24 went on to hepatic resection and 18 of them actually had a resection with curative intent. The median DFS and overall survival in the 24 patients with resection were 12 months and 28 months, respectively. When this was compared with the 6-month overall median survival in the 899 patients with hepatic metastases treated nonoperatively, it was significantly better. The two major contributors to improved survival seemed to be longer disease-free interval between the primary melanoma and the development of hepatic metastases, and limited hepatic disease that could be completely removed. For the select patients who would fit these criteria—and obviously there are not many—they seemed to benefit from hepatic resection.[64]

Unlike these other primary tumors, the data on the resection of colorectal metastases to the liver has been advancing exponentially over the last 30 years. In the United States, there are more than 50,000 patients each year with liver metastases from CRC. The rate of resection of these metastases has been increasing because of the expanding patient eligibility criteria. In the last decade there have been several retrospective reviews of resection of hepatic metastases from colorectal primaries—some with more than 1000 patients—that have added to the knowledge of which patients will benefit from resection. [54] [65] [66] [67] [68] [69] In the two largest series, the 5-year survival for patients with one to three metastases who had a resection was 30% or greater. [65] [66] Because of the strong feeling over the last three decades that resection was the optimal treatment for patients with one to three metastases, no randomized study of resection versus any other treatment has been performed. Two early studies compared the survival of matched historical control patients who underwent resection with those who had solitary hepatic lesions but did not undergo surgery. [70] [71] In both studies, not 1 of the 120 patients without resection survived for more than 3 years, whereas 30% of the resected patients survived for 5 years, underscoring the rationale for resection of solitary lesions.

The retrospective series from MSKCC reviewed their experience with 1001 liver resections in patients with colorectal metastases from the years 1985 to 1998.[65] Because the study was reported in early 1999, there were many patients who did not have a 5-year followup; in fact; the median followup of survivors was 32 months, which might not be long enough to tell us which patients would survive to 5 years. The median number of liver tumors was two, with 517 patients having solitary lesions and 330 having two or three lesions ( Table 59-2 ). The operative mortality was 2.8%. The 3-, 4-, and 5-year survival rates were 89%, 57%, and 37%, respectively; however, only 24.6% of patients resected before 1994 are 5-year survivors. The number of tumors removed (one or greater than one), the size of the tumors removed (greater or less than 5 cm), the preoperative CEA level (greater or less than 200 ng/mL), the extent of resection (less than or greater than a lobectomy), the resection margin in the hepatic specimen (negative or positive), and the presence of extrahepatic disease all were highly significant univariate and multivariate predictors of postsurgical survival. From these data, a clinical risk score was devised using five clinical criteria: the nodal status of the primary CRC, the disease-free interval from the primary CRC to the development of liver metastases, the number of hepatic tumors, the prehepatic resection CEA level, and the size of the hepatic tumors. Each criterion was given 1 point if the inferior condition existed, and then the points were added to give the score. The 5-year actuarial survival for a clinical risk score of zero was 60% as compared with a 14% survival for a score of 5. This gives surgeons a good insight into the prognostic expectations, but the score was not really intended for exclusion of patients from resection.


Table 59-2   -- Five-Year Survival after Hepatic Resection of Colorectal Cancer Metastases Based on Number of Lesions Resected

 

 

 

NO. OF METASTASES (% OF 5-YEAR SURVIVAL)

Study Group

No. of Patients

1

2–3

>1

>3

Memorial Sloan Kettering[65]

441

44

 

 

 

 

510

 

 

28

23

France[63]

1350

30

 

 

 

 

183

 

30

 

14

Mayo Clinic[68]

187

31

 

 

 

 

70

 

 

 

 

 

23

 

29

 

9

Italy[72]

134

20

 

 

 

 

78

 

 

17

 

Hepatic Registry[73]

789

37

37

 

18

Liver Met Survey[49]

2122

 

42[*]

 

26

*

<3.

 

Another large series was a multi-institutional report from France reviewing 1568 patients who underwent resection of liver metastases from CRC ( Table 59-3 ). [66] [72] Like the previous report, this study also looked at the effect of numerous prognostic indicators on overall survival after liver resection and then combined seven indicators into a prognostic scoring system. In this study, the 5-year survival among patients who had a resection of four or more lesions was 14% as compared with 30% for three or fewer nodules (P = 0.001). Both large studies, together with others, emphasize that the resection of more than four hepatic lesions results in significantly fewer cures, whereas for patients with one to three metastases, agreement exists that resection is worthwhile and can offer at least a 30% 5-year survival. [65] [66] [67] [68] [69] [70] [72] [73] [74] [75] [76]


Table 59-3   -- Prognostic Variables in the Multi-Institutional Study from France

Prognostic Values

5-Year Survival (%)

P Value

Age

 

 

 < 60 yr

28

0.06

 >60 yr

27

 

Tumor size

 

 

 <5 cm

30

0.002

 >5 cm

26

 

Stage of primary

 

 

 Dukes B

35

0.001

 Dukes C

21

 

Disease-free interval

 

 

 <2 years

26

0.002

 >2 years

32

 

No. of nodules resected

 

 

 <4

30

0.0001

 >3

14

 

Resection margin

 

 

 >1 cm

32

0.0006

 <1 cm

16

 

CEA

 

 

 <5

70[*]

0.0001

 >30

56[*]

 

Data from Nordlinger B, Guigiet M, Vallant JC, et al: Surgical resection of colorectal carcinoma metastases to the liver. Cancer 1996;77:1254–1262.

CEA, carcinoembryonic antigen.

 

*

2-year survival (%).

 

 

The French series found that one of the most significant prognostic variables was the stage of the primary lesion. For patients with CRC and negative lymph nodes (stage II), the 5-year survival after hepatic resection was 35% as compared with 26% for patients with mesenteric lymph node involvement (P < 0.001). Other studies addressing the issue of stage are listed in Table 59-4 . [73] [76] [77] [78] [79] [80] In the French study, preoperative serum carcinoembryonic antigen (CEA) was also significant; patients with a CEA value of less than 5 had a 2-year survival of 70%, as compared with a 56% 2-year survival for patients whose CEA was above 30. For patients with tumor nodules smaller than 5 cm, survival was 30% as opposed to 26% for larger nodules (P = 0.002). The age of the patient was not found to be significant in this study or others.[81] The time between the primary tumor and the development of liver metastases in this and other studies was significant for prognosis if the time periods were divided to include at least the first year with the synchronous lesions ( Table 59-5 ). [76] [77] [78] [79] [80]


Table 59-4   -- Survival after Hepatic Resection Based on Stage of Primary Colorectal Cancer

Study Group

No. of Patients

Dukes Stage

No. of Patients

5-Year Survival (%)

P Value

Milan[77]

95

B

29

47

0.02

 

 

C

50

24

 

United States[73]

789

B

226

47

0.001

 

 

C

317

23

 

Paris[78]

97

B

39

52

NS

 

 

C

58

53

 

Rotterdam[79]

117

B

53

26

NS

 

 

C

54

11

 

Erlangen[319]

173

B

49

53

0.01

 

 

C

119

32

 

Italy[76]

212

B

69

32

0.001

 

 

C

116

11

 

 

 


Table 59-5   -- Survival after Hepatic Resection Based on Synchronicity of the Hepatic Lesions and the Primary Colorectal Cancer

Study Group

No. of Patients

Synchronous

No. of Patients (in each group)

5-Year Survival (%)

P Value

Milan[77]

95

Yes

28

31

NS

 

 

No

67

28

 

Rotterdam[79]

117

Yes

38

15

NS

 

 

No

79

22

 

Erlangen[319]

173

Yes

85

32

0.05

 

 

No

88

45

 

Heidelberg[320]

122

Yes

48

2

0.01

 

 

No

74

16

 

United States[73]

789

Yes

259

27

 

 

 

1–12 mo

206

31

 

 

 

>12 mo

333

42

0.02

Paris[78]

97

Yes

35

52

NS

 

 

No

62

52

 

Italy[76]

212

Yes

85

18

NS

 

 

No

113

22

 

 

 

There have been conflicting reports about the importance of the margin of resection when removing hepatic metastases. Although the cutoff used in the French study was different from that used in the Memorial study, the margin of normal hepatic tissue around the resected metastases was significant. For those patients with a margin greater than 1 cm, 32% of patients survived for 5 years, as opposed to a 16% survival for those with a margin less than 1 cm (P = 0.006). A report from Germany in 1991, however, showed no significant difference in survival among over 170 patients relative to their margins of resection.[80] This study assigned the margins to three groups: 1–4 mm, 5–9 mm, and greater than 10 mm. More recent studies support this study by showing that if there is no margin of clearance between the tumor and normal liver parenchyma, patients do worse, but whenever there is a margin, the difference between 1 mm, 10 mm, or greater is not significant ( Table 59-6 ). [40] [41] These data are understandable from a pathologic standpoint, because these tumors tend to be firm nodules with pushing borders and even a small margin would remove all of the tumor, whereas a positive margin would leave cells behind. This would not be true for primary hepatocellular carcinomas (HCCs), as they can have more diffuse, infiltrative borders.


Table 59-6   -- Significance of Hepatic Resection Margin

Institution

Margin of Resection

No. of Patients

5-Year Survival (%)

P Value

Erlangen[80]

0–4 mm

67

23

NS

 

5–9 mm

40

29

 

 

>10 mm

65

39

 

Mayo[69]

0–1 mm

17

29

NS

 

1–10 mm

123

30

 

 

>10 mm

31

36

 

 

None

24

17

 

Pittsburgh[68]

0–10 mm

92

25

 

 

>10 mm

95

29

 

 

None

17

0

0.006

Memorial[321]

1–10 mm

248

43

 

 

>10 mm

113

43

 

 

None

65

17

0.00003

 

 

The French authors created a scoring system by using these seven variables, giving one point for each bad variable and two points for a CEA value over 30.[66] A 2-year survival of 79% was calculated for patients with a score of 0 to 2 points, 60% if they had scores of 3 to 4 points, and 43% if they scored 5 to 7 points.[66] These prognostic scoring systems from the French study and the Memorial study are valuable tools to estimate patients’ survival after a resection. For patients with poorer prognoses, other therapies might be added to surgery to try to improve their chances. [66] [72]

Although recent studies have suggested that resection of four or more lesions can result in 5-year survivals equivalent to those after resection of one to three lesions, these reports must be examined with caution. A report of 98 patients with hepatic resection of four or more colorectal metastases reported an actuarial survival of 33%. Yet the median followup was only 33 months, which is too short for patients with liver resections. The DFS for these patients was 12 months, and 83% of the patients had recurrence. Of this whole group there were only seven actual 5-year survivors, and all of them have recurrent disease. One important point from this article is that the worst survival was seen in patients whose disease was progressing while receiving neoadjuvant therapy before the liver resection.[82]

The Liver Met Survey is an international Internet-based registry that has entered 2122 patients. The 5- and 10-year survivals are 42% and 26%. The 5-year survivals for three or fewer nodules and more than three nodules were 48% and 24% (P = 0.0001). Another variable that affected survival concerned whether the tumor was unilateral or bilateral. In this review, preoperative chemotherapy did not benefit patients with solitary metastases (5-year survival of 45% vs. 58%), whereas for patients with more than five metastases, 5-year survivals were 22% and 12% for patients with and without previous chemotherapy.[83]

A group from Paris reporting on surgery for patients who had been downstaged by chemotherapy also found that if the tumors’ pathologic response to chemotherapy was less than complete necrosis, the survival after resection was significantly decreased (P = 0.002).[84] This group of patients with more than four lesions and progressing disease should probably not be candidates for surgical resection. The use of the PET scan may help increase the survival of patients with liver resections from colorectal primaries because of improved selection of patients. A study of 100 patients who were shown to have disease confined to their liver by PET scan reported a 5-year survival of 58%. PET scan is believed to detect unsuspected tumors in 25% of patients who would have otherwise been considered to have only resectable hepatic metastases. This was not a randomized study, so it is difficult to determine if the PET scan alone was responsible for this improved survival over the usual 30% 5-year survival. However, it may be prudent at this time to perform a PET scan before taking patients to liver resection, so that those who have incurable disease will be spared major surgery.[85]

The issue of trying to downstage patients with unresectable liver metastases and then proceed to liver resection is becoming more relevant as newer and more effective chemotherapy agents are being introduced. A report from Paris followed 1104 patients with liver metastases from colorectal primaries considered to be unresectable. Of these patients, 138 (13%) were downstaged sufficiently to go on to surgery. Eighty percent of these patient developed tumor recurrence, 72% of which involved the liver. DFS was 17% at 5 years. These numbers reflect that the overwhelming majority of patients will not be downstaged by chemotherapy sufficiently for resection, and even those that are do not fare particularly well.[84]

Several studies have addressed the issue of extrahepatic intraabdominal metastases at the time of hepatic resection. In most instances, surgeons do not proceed with liver resection in the presence of extrahepatic metastases. In the series from New York, 88 patients with extrahepatic disease were included.[19] More than half of these patients had direct extension into other organs such as the diaphragm. Only 10 patients had positive portal nodal disease. Looking at all 88 patients with extrahepatic disease, the 5-year actuarial survival was 18%. The breakdown between those with discontinuous disease or direct extension was not made. In the Mayo Clinic report on hepatic resections in patients with extrahepatic disease, none of the 22 patients survived for 5 years, and only 1 survived for 3 years.[86] In the Hepatic Registry, of the 61 patients with extrahepatic involvement undergoing hepatic resection, none had a 5-year DFS.[74] These data support the view that hepatic resection in the presence of noncontiguous extrahepatic intraabdominal disease, with the exception of a local recurrence, is rarely curative and, in general, is inadvisable.

For extrahepatic disease not in the abdomen, the usefulness of resection could be different for disease in the lungs, especially solitary lesions in the lung. A recent study from the Mayo Clinic reviewed their experience with resection of both hepatic and pulmonary metastases from colorectal primaries. There were 58 patients, with no operative mortalities and a 5-year survival rate of 30%. These authors believed that the resection of both lung and liver of selected cases was justified.[87]

Relapse rates after liver resections are described in Table 59-7 . In the larger studies it seems that approximately 40% of patients who have hepatic resections will have recurrent disease in the liver as the first sign of the relapse. Of the 69 patients who had relapse in the Milan study, 28 (41%) had relapse in the liver only, 19 (28%) had only extra-abdominal relapse, 9 (13%) had intraabdominal extrahepatic relapse, and 13 (19%) had relapse in both the liver and at an extrahepatic site.[49] Although extrahepatic failure is of concern, the liver remains the main site of relapse, appearing in more than 60% of patients.


Table 59-7   -- Recurrence after Hepatic Resection

Author

No. of Patients

No. of Recurrences

No. of Liver Only

No. of Liver and Extrahepatic

No. of Extrahepatic Only

Van Ooijen et al[79]

117

69

20 (29%)

14 (21%)

34 (50%)

Codi et al[77]

93

69

28 (41%)

13 (19%)

28 (41%)

Hohenberger et al[320]

122

80

17 (21%)

55 (69%)

8 (10%)

Hughes et al[75]

607

424

148 (27%)

154 (27%)

106 (28%)

Rees et al[322]

89

61

25 (41%)

9 (15%)

27 (44%)

Fong et al[321]

465

235

96 (41%)

16 (7%)

123 (52%)

 

 

Because of the increased use of hepatic resection for metastatic liver disease, the incidence of repeat hepatic resection has also increased. Approximately 10% of patients who have had a hepatic resection can have a repeat resection ( Table 59-8 ). A study from Paris of 116 patients who underwent repeat hepatic resection reported a low operative mortality of 0.9% and a 3-year survival of 33%.[54] Of the patients who underwent repeat hepatic resection, 55% had recurrence in their liver after this operation. There were 170 patients in a registry report who had repeat hepatic resections with a 5-year survival of 26%, which was comparable to the 5-year survival for the original hepatic resection.[88] A recent study from MSKCC reviewed 126 second liver resections for recurrent colorectal metastases; the 5-year actuarial survival was 34%, with 19 actual 5-year survivors. The operative mortality was 1.6%, and morbidity was 28%.[89] A study of recurrence among these patients revealed a liver recurrence in 67% of the patients. Repeat hepatic resection, when feasible, can be done safely, and the outcome is comparable to that for the original hepatic resection.[90] Thus, if patients have isolated liver metastases (preferably solitary lesions) after hepatic resection, they should be candidates for repeat resection.


Table 59-8   -- Survival after Repeat Hepatic Resections for Colorectal Metastases

 

 

 

SURVIVAL (%)

Author

No. of Patients

Operative Mortality (%)

2-Year

3-Year

5-Year

Nordlinger et al[54]

116

0.9

57

33

 

Petrowsky et al[89]

126

1.6

 

51

34

Fernandez-Trigo et al[88]

170

 

 

37

26

Que et al[323]

21

 

 

50

 

 

 

The place for ablation of liver metastases whether by radiofrequency (RF) heating or cryotherapy freezing has not yet been delineated adequately. Because extensive reporting on resection has indicated an expected 30% 5-year survival for patients with one to three hepatic metastases from colorectal primaries, the use of ablation in this setting cannot be warranted until a randomized study has shown that the two therapies are equivalent. A retrospective study from M.D. Anderson Cancer Center analyzed a series of 418 patients with liver-only metastases from colorectal primaries, of whom 57 had RF only. The overall recurrence rate and the liver recurrence rate were both significantly higher for the RF group over the resection-only group, with a liver recurrence of four times greater in the RF group. The overall survival and DFS was also significantly better in the resection group. A multivariate analysis continued to show RF as a significant factor in poor prognoses. Interestingly, 31 patients with solitary tumors were treated with RF (we are not told why this happened), and these patients had significantly worse survival than the patients with resection of solitary lesions (P = 0.025). The group of patients who had RF alone when compared with those 70 patients with chemotherapy alone had a significantly improved survival.[91] A subsequent report from the same group of investigators compared RF to resection for patients with solitary colorectal metastases. The hepatic recurrence rate was much higher in the RF group than in the resection group, 37% as compared with 5%, respectively. The 5-year recurrence-free survival and overall survival were 40% and 50% as compared with 0% and 0% for resection versus RF, respectively. The conclusions are obvious: resection is the preferred method for treatment of solitary metastases whenever possible.[92]

Synchronous Liver Metastases

The question of whether liver metastases can safely be resected at the time of resection of the primary colorectal carcinoma is still uncertain. Small metastases seen at the time of laparotomy for the primary lesion can easily be resected. When patients are evaluated for their primary and are found to have a large burden of metastatic disease, a question may be whether a simultaneous resection of the primary tumor and the metastases should be done, or whether preoperative chemotherapy should be offered. Bolton and Fuhrman reported a 12% operative mortality with simultaneous surgeries that increased to 24% if the surgery included a major liver resection.[93] Nordlinger reported a 7% mortality for simultaneous resections compared with 2% for staged resections (P = 0.01).[94] Tanaka and coworkers, Martin and colleagues, and Weber and associates, on the other hand, reported that operative mortality and morbidity were comparable whether staged or simultaneous procedures were done. [95] [96] [97] It is the practice in many institutions to do a simultaneous resection with right colon primaries or when single synchronous metastases are found in the liver, and staged resections for rectal primaries or for patients with multiple liver metastases.[98] Some advocate systemic chemotherapy first, followed by resection of liver first, and then resection of the colon[99] or simultaneous resection.[100]

Adjuvant Therapy after Liver Resection

Despite the high curative resection rate of hepatic resections, the recurrence of metastases is around 70%, with 50% recurring in the liver. The use of hepatic artery infusion (HAI) of chemotherapy after liver resection was studied in a small prospective randomized fashion at the City of Hope Medical Center. [101] [102] The patients with solitary metastases all had resection of their tumors, and half received postoperative continuous HAI of fluorodeoxyuridine (FUDR). For the six patients who had resection only, their median time to failure was 8.7 months, and three of the six metastases recurred in the liver. For the five patients with resection plus pump, none had recurrence in the liver, and their median time to failure was 30.7 months. There was no difference in median survival between the two groups.

At MSKCC, 156 patients were randomized after liver resection to either HAI with systemic chemotherapy (HAI + SYS) or systemic chemotherapy alone (SYS). Chemotherapy was administered for 6 months. The endpoint of this study was 2-year survival. HAI therapy used was FUDR and dexamethasone (Dex), and the systemic therapy was 5-fluorouracil (5-FU) and leucovorin (LV) or continuous infusion of 5-FU. Patients were stratified by type of previous chemotherapy or no chemotherapy and the number of liver metastases (one, two to four, more than four). Two-year survival was increased in the group receiving HAI + SYS (86%) versus 72% for SYS (P = 0.03).[103A] Median survival is presently 68.4 months for the HAI + SYS group and 58.8 months for the group receiving SYS alone. With a median 10-year followup, the 10-year survivals are 41% and 27%, respectively ( Fig. 59-5 ).[104] Hepatic DFS is clearly better for the HAI + SYS group, with a median survival not reached, as compared with 32.5 months with SYS alone (P = 0.003; Fig. 59-6 ). Overall, DFS is also increased significantly (31.3 and 17.2 months for the HAI + SYS vs. SYS alone [P = 0.02]; Fig. 59-7 ). Toxicity was increased in the combined group, with increased diarrhea and increased liver function test abnormalities. A total bilirubin greater than 3 mg/dL occurred in 18% of patients receiving HAI and in 2% of the SYS group.[103]

 
 

Figure 59-5  Survival curve. Combined modality treatment: HAI FUDR + systemic 5-FU/LV. Monotherapy: systemic 5-FU/LV.[104]

 

 

 
 

Figure 59-6  Hepatic progression-free survival. HAI (FUDR/dex) + SYS (5-FU/LV) versus SYS (5-FU/LV) alone. P < 0.0001.[104]

 

 

 
 

Figure 59-7  Progression-free survival. HAI (FUDR/dex) + SYS (5-FU/LV) versus SYS (5-FU/LV) alone. P < 0.02.[104]

 

 

The Eastern Cooperative Group (ECOG) and the Southwestern Oncology Group (SWOG) conducted a randomized study of hepatic resection alone versus resection followed by 4 cycles of HAI-FUDR and 12 cycles of systemic infusion of 5-FU. Only patients with three or fewer metastases were enrolled in the study. The study was powered to answer the question of whether HAI would increase DFS. Of the 109 patients randomized, only 75 are actually in the study, because several patients were excluded because of extrahepatic disease, unresectable disease, or no tumor. Four-year liver recurrence-free survival was 67% in the chemotherapy group and 43% in the control group (P = 0.03). Median survival was 63.7 months for the chemotherapy group and 49% for the control group.[105] The endpoint was obtained with a 4-year DFS of 46% with HAI + SYS and 25% for the control group (P = 0.035).

A German cooperative group enrolled patients from 26 different centers and entered 226 patients into a study of resection followed by adjuvant hepatic arterial therapy with 5-FU and LV via a port as compared with a control group receiving no chemotherapy after resection. Although 113 were entered into the hepatic arterial group, only 87 were treated. Only 64% had chemotherapy data available, and only 30% completed treatment. At an 18-month interim analysis, the relapse rate was 33% in the group receiving adjuvant therapy and 36% in the group treated with resection alone. If one looks only at those who were treated, the median survival was 44.8 months from the treated group versus 39.7 months in the control group.[106]

A randomized study from Greece used intraoperative randomization to regional mitomycin C, 5-FU/LV with interleukin-2 via a hepatic arterial catheter with the same drugs via a systemic route versus systemic therapy alone. The 2-year survivals in 122 patients were 80% and 71%, and 5-year survivals were 73% and 60%, in the regional-plus-systemic versus systemic-alone groups, respectively (P = 0.004).[107] Five-year hepatic-free recurrence was also significantly increased in the regional-plus-systemic group at 82% versus 49% in the systemic-alone group (P < 0.001). Table 59-9 summarizes survival data of the randomized studies and the disease-free survival from these studies.


Table 59-9   -- Randomized Trials of Adjuvant Therapy and Studies after Liver Resection: HAI versus SYS or Control

 

 

2-YEAR SURVIVAL (%)

5-YEAR SURVIVAL (%)

Author

No. of Patients

HAI

SYS or Control

HAI

SYS or Control

Kusonoki et al[110]

58

80

71

73

60

MSKCC[103]

156

86

72

57

48

Lorenz et al[106]

201[*]

60

62

50

30

ECOG[105]

75[†]

70

65

60

35[‡]

Lygidakis et al[107]

122

92

75

73

60

Asahara et al[111]

38

100

60

100

47[§]

 

 

DISEASE-FREE SURVIVAL

 

 

2-YEAR (%)

 

5-YEAR (%)

 

Studies

No. of Patients

HAI

SYS

 

HAI

SYS

P Value

MSKCC[104]

156

55

40

 

40

30

0.02

ECOG[105]

75

60

40

 

40

20[#]

0.03

Lorenz et al[106]

186

 

 

Median 20/12.6[#]

 

 

NS

Lygidakis et al[107]

122

66

48

 

60

35

0.0002

Tono et al[108]

19

75

30

 

60

20

0.045

*

Treated patients, not everyone randomized.

Patients entered in study, not everyone randomized.

Updated figures.[107]

§

4-year survival.

#

No treatment in control arm.

 

Tono and coworkers[108] randomized 19 patients to continuous infusion of 5-FU, 500 mg per day for 4 days via HAI for 6 weeks. Although this is a small study, there was an increase in 3-year DFS, 66.7% for the regional group and 20% for the control group (P = 0.045). The 5-year survivals were 77.8% for the regional group versus 50% for the control group.

In a nonrandomized study comparing two groups of patients in Japan,[109] the survival was increased with the use of adjuvant regional therapy after hepatectomy for colon cancer. The 5-year survivals were 23% using surgery alone, and 57% when adjuvant chemotherapy with 5-FU, doxorubicin, and mitomycin C were given after surgery. In another nonrandomized study from Japan,[110] 58 patients who had radical resection of metastatic colorectal carcinoma could select whether they wanted HAI + systemic, or systemic alone after surgery. The 5-year survival was significantly increased for the HAI + systemic group at 59%, as compared with 27% for the systemic-alone group (P < 0.001), as was hepatic recurrence at 7% versus 57% (P < 0.001). In another small study, 4-year survival was 100% as compared with 47% for the HAI and control groups, respectively (P = 0.05).[111]

Studies are now testing the new systemic chemotherapeutic agents with HAI as adjuvant therapy after liver resection. A phase I study at MSKCC with systemic CPT-11 (irinotecan) and HAI FUDR/Dex had a 2-year survival of 89% and a 5-year survival of 58%.[112] Another trial using escalating systemic doses of oxaliplatin (Oxali)/5-FU/LV and HAI therapy,[113] with a minimum followup time of 26.9 months, had a 2-year survival of 98%. At the Mayo Clinic, FUDR + Dex has been added to systemic Oxali and capecitabine. Presently, the 2-year survival is 86%.[114]

Ongoing studies to address the usefulness of adjuvant therapy after liver resection include the National Surgical Adjuvant Breast and Bowel Project study of HAI + systemic Oxali + capecitabine versus systemic Oxali + capecitabine alone. A trial at MSKCC is addressing the question of the safety of bevacizumab (Bev) given with HAI + systemic therapy after liver resection in a randomized study of HAI and systemic therapy with or without bevacizumab. Few studies have evaluated the utility of systemic therapy after liver resection. In the ENG study (EORTC/NCIC CTG/GIVIO), 128 patients were randomized to receive 5-FU and LV chemotherapy for 8 weeks versus no further therapy (control) after liver or lung resection. There were no significant differences in DFS or survival between the two groups. The 4-year survivals were 57% and 47% for the treated and control groups.[115] A European intergroup study (FFCD) randomized 173 patients to systemic 5-FU/LV versus no further treatment. The 5-FU and LV were given by the bolus method on a monthly schedule. Two-year DFS was 50.4% for those receiving chemotherapy and 38.1% for the control group (P = 0.058). Negative prognostic factors were synchronous disease, multiple metastases, stage III tumor, preoperative hypertension, postoperative complications, and an elevated preoperative CEA. Two- and five-year survivals were 81% and 51% for the chemotherapy group and 82% and 41% for the control group. A trial combining these two studies to increase power demonstrated a DFS of 27.9 versus 18.8 months (P = 0.058) and overall survival of 62.2 versus 47.3 months (P = 0.095) for the chemotherapy versus surgery-alone groups, respectively. Patient characteristics in this study included one metastasis in 68% of patients, and more than 1 year disease-free interval in 57% of patients. This clearly can affect results, as seen in their own analysis, where DFS was 27 months for one metastasis and 16.8 months for two or more metastases (P = 0.036), and survivals were 64.5 and 40 months for one or more than two metastases, respectively. In the MSKCC adjuvant therapy of HAI + systemic 5-FU/LV after liver resection study, the patients had worse baseline characteristics, with 36% of patients having only one metastasis, whereas 19% had more than four metastases, and only 20% had a disease-free interval longer than 1 year.[103]

The efficacy of systemic CPT-11 in posthepatic resection was addressed in 29 patients of whom 62% had only one metastasis, 79% had a disease-free interval longer than 1 year, and 62% had a surgical margin larger than 1 cm. Their median DFS was 45 months, and 2-year survival was 85%.

Adjuvant therapy following resection has also been attempted in HCC. In patients with HCC, a randomized study compared hepatic arterial injection of epirubicin and oral 1-hexycarbamoyl-5-fluorouracil (HCFU) to no further treatment in 57 patients after liver resection of HCC.[116] There was no significant difference between the two groups, and the authors believed that the chemotherapeutic agents chosen might not have been appropriate. Postoperative adjuvant therapy could compromise survival if the treatment is not effective, because it suppresses host immunity. In a small randomized study of intravenous epirubicin with hepatic arterial iodized oil and cisplatin compared to no further treatment after liver resection for HCC, the DFS was shorter in the treated group. There was an increase in extrahepatic metastases in the treated group.[117]

Factors that inhibit hepatic carcinogenesis might be more useful as adjuvant therapies for HCC. Retinoic acid inhibits chemically induced hepatocarcinogenesis in rats, spontaneous HCC in mice, and the production of α-fetoprotein in human hepatoma cell lines. In a study by the Hepatoma Prevention Study Group, 89 patients were randomized to polyphenolic acid (acyclic retinoid) or placebo for 12 months after liver resection. The polyphenolic group had a significant reduction in recurrent or new hepatomas (P = 0.04). The drug also improved overall survival, but not significantly.[118]

SYSTEMIC CHEMOTHERAPY

Responses of liver metastases to systemic chemotherapy are variable but usually reflect the response of the primary tumor. Most studies of systemic therapy do not differentiate patients who have only liver metastases, making it difficult to draw conclusions about the usefulness of systemic chemotherapy to treat liver metastases. In some cases, however, the response rates of liver metastases are well documented.

In breast cancer, liver metastases represent a poor prognostic indicator, with median survival of 10 months in some series.[119] Visceral metastases, including liver metastases, have been reported to have fewer estrogen-positive receptors. Insulin-like growth factors are present in the liver and the lung and could be important to the growth and motility factors for breast cancer and lung cancer.[120] In looking at prognostic factors that predict response, patients with liver metastases were the ones who tended to respond the least.[121] Carter's[122] review of single agents (5-FU or cyclophosphamide) demonstrated a 20% response in liver metastases vs. 32% and 27% in soft-tissue and osseous metastases, respectively. The combination of halotestin, prednisone, and 5-FU produced objective responses in 26 of 52 patients (50%) with liver metastases from breast carcinoma.[84] Although combination chemotherapy has substantially improved the response rates obtained in treating breast cancer, liver metastases still have a lower response rate than soft-tissue or pulmonary disease. In a Southwest Oncology Group study of 262 patients, 41 of 88 patients (47%) with liver metastases responded, as compared with 110 of 154 patients (71%) without liver involvement (P = 0.001).[123] There is a suggestion that taxoids might be more effective against liver metastases.[124]

For metastases from gastric carcinoma, one of the more commonly used regimens (5-FU, doxorubicin, and mitomycin C) produces a mean response of 35%, whereas for those with hepatic metastasis, the response rate was 28%.[125] The cumulative response for liver metastases was 34%, while the overall mean response rate was 36% for combination chemotherapy in the treatment of gastric carcinoma.[126]

For patients with CRC, the liver is the most common site of dissemination, with as many as 70% of patients with metastatic disease developing liver metastases.[127] Several new agents are now available, among which are irinotecan (CPT-11) and oxaliplatin (Oxali). [129] [130] Randomized trials using CPT-11 with 5-FU/LV versus 5-FU/LV alone [131] [132] produced an increase in response rate and survival. When CPT-11/5-FU/LV (IFL) was compared to Oxali + 5-FU/LV (FOLFOX), the response rate was increased from 35% to 45%, and the survival was increased from 15 to 19.5 months for the IFL and FOLFOX groups, respectively.[132] When 5-FU was changed to infusion FOLFIRI (CPT-11/5-FU/LV) it became more effective and produced results that are similar to FOLFOX.[133] In the last few years, targeted agents have become available: Bev, a monoclonal antibody to vascular endothelial growth factor and cetuximab (C225), an antibody to epidermal growth factor receptor. The addition of Bev to IFL[134] increased response rates and survival (35% to 45%, and 15.6 to 19.5 months, for IFL vs. Bev + IFL, respectively). For almost four decades, the 2-year survival for metastatic colorectal patients treated with 5-FU or 5-FU + LV was 25%; with these new agents, 2-year survival has moved up to 30% to 39%, with a marked improvement in overall survival from 12 to 20 months.

When does one start systemic chemotherapy for patients with metastatic disease to the liver from breast, gastric, or colon cancer? For patients with colon or gastric cancer who are asymptomatic and have a small volume of disease (<20% of the liver involved with tumor and not resectable), three studies suggest that earlier treatment increases survival and increases the time with good performance status. [136] [137] [138] If the patient has rapidly progressive disease or is symptomatic, chemotherapy should be initiated immediately. One should have a baseline CT scan and laboratory values before starting treatment. To assess response, scans should be repeated at 2- to 3-month intervals to evaluate whether the tumor is responding, and therapy should be continued.

Hepatic Arterial Chemotherapy

The rationale for hepatic arterial chemotherapy has an anatomic and pharmacologic basis ( Table 59-10 ):

  

1.   

Liver metastases are perfused almost exclusively by the hepatic artery, whereas normal hepatocytes derive their blood supply both from the portal vein and minimally from the hepatic artery.[138]

  

2.   

Certain drugs are mostly extracted by the liver during the first pass through the arterial circulation. This results in high local concentrations of drug with minimal systemic toxicity. Ensminger and coworkers [140] [141] demonstrated that 94% to 99% of FUDR is extracted by the liver during the first pass, compared with 19% to 55% of 5-FU. This makes FUDR an optimal drug for hepatic arterial chemotherapy. The pharmacologic advantages of various chemotherapeutic agents for HAI are summarized in Table 59-11 . Newer drugs, such as CPT-11 and etoposide, do not show an advantage when given by HAI.[141]

  

3.   

Drugs with a steep dose-response curve will be more useful when given by the intrahepatic route, because a large dose can be given regionally.

  

4.   

Drugs with a high total-body clearance are more useful for hepatic infusion. If a drug is not cleared rapidly, recirculation through the systemic circulation diminishes the advantage of hepatic arterial delivery.[142]

  

5.   

The liver is often the first and only site of metastatic disease. The theory of the stepwise pattern of metastatic progression states that hematogenous spread occurs first via the portal vein to the liver, then from the liver to the lungs, and then to other organs. [144] [145] Thus aggressive treatment of metastases confined to the liver (resection and/or hepatic infusion) could yield prolonged survival for some patients.


Table 59-10   -- Rationale for Hepatic Artery Infusion[152]

  

   

Liver metastases are perfused by the hepatic artery. Normal liver is primarily supplied by the portal vein.

  

   

Certain drugs have high hepatic extraction.

  

   

The liver is often the first site of metastases; eliminating liver metastases would prevent extrahepatic disease.

  

   

Many drugs have a steep dose-response curve.

  

   

Drugs with a high total-body clearance are more effective.

 

 


Table 59-11   -- Drugs for Hepatic Arterial Infusion

Drug

Estimated Half-life (min)

Increase on Exposure by HAI

5-FU

10

5- to 10-fold

FUDR

<10

100- to 400-fold

BCNU

<5

6- to 7-fold

Mit C

<10

6- to 8-fold

Cisplatin

20–30

4- to 7-fold

Adriamycin (doxorubicin hydrochloride)

60

2-fold

BCNU, bischloroethyl nitrosourea or carmustine; 5-FU, fluorouracil; FUDR, 5-fluoro-2-deoxyuridine; HAI, hepatic arterial infusion; Mit C, mitomycin C.

 

 

 

Regional hepatic arterial therapy can be done by using either a hepatic arterial port or a percutaneously placed catheter connected to an external pump or to a totally implantable pump. Early studies with percutaneously placed hepatic artery catheters produced high response rates, but clotting of the catheters and the hepatic artery, as well as bleeding, led physicians to abandon this method ( Table 59-12 ). The development of a totally implantable pump allowed long-term HAI with good patency of the catheter and the hepatic artery and a low incidence of infection. One study compared three groups relative to placement of the hepatic artery catheter:

  

1.   

Surgical placement of a hepatic artery catheter

  

2.   

Percutaneous placement of the hepatic artery catheter

  

3.   

An operative implantable reservoir connected to the hepatic artery catheter


Table 59-12   -- Hepatic Artery Infusion with External Pump

Investigators

No. of Patients

Drug

Response (%)

Catheter Complications or Bleed (%)

Tandon et al[324]

122

5-FU 25 mg/kg × 9 days

65

42

Ansfield et al[325]

419

5-FU 25 mg/kg × 4 days

55

21

Watkins et al[326]

184

5-FU 25 mg/kg × 10 days

71

28

Cady and Oberfield[327]

55

FUDR 20 mg/kg

67

39

Smiley et al[328]

166

5-FU 25 mg/kg × 4 days

25

30

New studies

 

 

 

 

 Metzger et al[160]

30

5-FU 2 g ×5 days

57

33

 Denck[329]

50

5-FU 6 g × 3 days

58

5

 Schlag et al[330]

33

5-FU 1 g × 5 days

27

20

 Rougier et al[331]

43

5-FU 1 g weekly

56

65

 Arai et al[332]

32

FU 1000 × 5 hrs every week

78

25

 

 

Mit C 10–12 mg/m2 every 6 weeks

 

 

5-FU, 5-fluorouracil; FUDR, 5-fluoro-2-deoxyuridine; Mit C, mitomycin C.

 

 

 

The reported ability of each technique to administer chemotherapy for the three groups of patients was 31, 25, and 115 days, respectively.[145] Several trials using the implantable pump produced high response rates with good survivals ( Table 59-13 ).


Table 59-13   -- Hepatic Arterial FUDR Infusion with Internal Pump: Responses

Investigator

No. of Patients

Prior Chemotherapy (%)

PR (%)

Decrease in CEA (%)

Median Survival (mo)

Niederhuber et al[333]

70

45

83

91

25

Balch and Urist[334]

50

40

83

26

Kemeny et al[335]

41

43

42

51

12

Shepard et al[336]

53

42

32

17

Cohen et al[337]

50

36

51

Weiss et al[338]

17

85

29

57

13

Schwartz et al[339]

23

15

75

18

Johnson et al[340]

40

47

12

CEA, carcinoembryonic antigen; FUDR, 5-fluoro-2-deoxyuridine; PR, partial response.

 

 

 

Randomized Studies

It is difficult to assess the impact of hepatic infusional therapy on tumor response and patient survival without a prospective randomized study comparing it with systemic chemotherapy. In such studies, patients would have to be stratified for parameters known to influence tumor response rates and patient survivals, such as performance status, extent of liver involvement, and initial LDH level.

The influence of the extent of liver involvement on survival has been demonstrated by many investigators. The median survival for patients with less than 20% involvement was greater than 20 months, whereas survival was only 6 months for those with more than 60% involvement ( Fig. 59-8 ).[146] The influence of certain laboratory parameters on tumor response and patient survival was evaluated at MSKCC, where the initial LDH level proved to be the most significant factor.[147] Patients whose initial LDH and CEA levels were normal had a median survival of 32 months as compared with only 8 months for those with abnormal values ( Fig. 59-9 ).

 
 

Figure 59-8  Survival distributions by percentage of liver involvement (medical assessment). The median survival was 25 months for patients with less than 20% involvement vs. 6 months for patients with greater than 60% involvement.  (From Kemeny N, Daly J, Oderman P, et al: Prognostic variables in patients with hepatic metastases from colorectal cancer: importance of medical assessment of liver involvement. Cancer 1989;63:742.)

 

 

 

 
 

Figure 59-9  Metastatic colorectal carcinoma (survival). Survival curves based on initial LDH and CEA levels. The median survival of patients with normal LDH and CEA levels at initiation of chemotherapy was 32 months, whereas it was 8 months if both values were abnormal (P < 0.001).  (From Kemeny N, Braun DW: Prognostic factors in advanced colorectal carcinoma: the importance of lactic dehydrogenase, performance status, and white blood cell count. Am J Med 1983;74:786.)

 

 

 

In a randomized study at MSKCC, patients were stratified by percentage of liver involvement and baseline LDH. All patients underwent exploratory laparotomy for pump placement; patients with extrahepatic disease were excluded from entry. Both groups received a 14-day continuous infusion of FUDR, but the dose was lower in the systemic group than in the HAI group. In patients randomized to systemic therapy, crossover from systemic therapy to HAI was permitted. Of the 99 evaluable patients, partial responses were seen in 53% and 21% of the HAI and systemic groups, respectively (P = 0.001). Of the patients who crossed over from systemic to HAI therapy, 25% had a partial response after the crossover and 60% had a decrease in CEA levels. The median survivals for the HAI and systemic groups were 17 and 12 months, respectively (P = 0.424). The interpretation of survival is difficult, because 60% of the systemic patients crossed over. Those who did not cross over had a median survival of 8 months, as compared with 18 months for those who crossed over to HAI (P = 0.04). Randomized studies are outlined in Table 59-14 .


Table 59-14   -- Randomized Studies of Intrahepatic versus Systemic Chemotherapy for Hepatic Metastases from Colorectal Cancer

 

 

RESPONSE (%)

 

SURVIVAL (MO)

 

Group

No. of Patients

HAI

SYS

P Value[*]

HAI

SYS

P Value[*]

MSKCC[341]

162

52

20

0.001

18[*]

12

 

NCOG[342]

143

42

10

0.0001

16.6

16

 

NCI[148]

64

62

17

0.003

20

11

 

Consortium[343]

43

58

38

NS

NS

 

City of Hope[102]

41

56

0

 

 

 

 

Mayo Clinic[344]

69

48

21

0.02

12.6

10.5

 

French[345]

163

49

14

NS

15

11

0.02

English[135]

100

50

0

0.001

13

6.3

0.03

German[150]

168

43

20

0.019

12.7

18

 

CALGB[152]

134

43

27

0.02

24.4

20

0.0034

HAI, hepatic artery infusion; NS, not stated; SYS, systemic chemotherapy.

 

*

Only significant P values reported.

 

A similar randomized study conducted by the Northern California Oncology Group also used FUDR infusion in both the HAI and systemic groups. A total of 114 patients were stratified by extent of liver involvement, baseline bilirubin values, and performance status. Of the 117 eligible patients, 42% responded to HAI and 10% to systemic therapy (P < 0.001). The median times to progression were 401 and 201 days for the HAI and systemic groups, respectively (P = 0.009), whereas median survivals were 503 days and 484 days, respectively. Although a crossover design was not built into the study, 43% of the patients on systemic therapy eventually received HAI, which could have obscured survival differences between the two groups.

A National Cancer Institute (NCI) study compared HAI with systemic infusion of FUDR.[148] In 64 patients, the response rates were 62% and 17% for the HAI and systemic groups, respectively (P < 0.003). Interpretation of survival data is difficult, because 34% of the HAI group never received chemotherapy, and 38% of the HAI group had positive portal lymph nodes. The 2-year survival was 47% in the HAI group compared with 13% in the systemic group (P = 0.03).

Another small study conducted by the Mayo Clinic (69 patients) compared HAI FUDR with systemic bolus 5-FU for 5 days.[123] Objective tumor response was observed in 48% and 21% of patients in the HAI and systemic groups, respectively (P = 0.02), and times to hepatic progression were 15.7 and 6 months, respectively (P = 0.001). Survival was similar in the two groups (12.6 and 10.5 months, respectively). Forty-eight percent of the HAI groups either were not adequately treated or had extrahepatic disease.

In a French trial of 163 patients randomized to HAI of FUDR versus systemic-bolus 5-FU, the patients were stratified by extent of liver involvement and baseline LDH levels.[118] The response rates were 49% and 14% in the HAI and systemic groups, respectively. Median times to hepatic progression were 15 and 6 months, and median survivals were 14 and 10 months for the HAI and systemic groups, respectively. The 2-year survival was 22% for HAI and 10% for the systemic group (P < 0.02).

In a similar study done in England, 100 patients were randomized to HAI FUDR or systemic 5-FU (which was given to symptomatic patients only).[149] Quality of life and survival were improved significantly for the HAI group. Median survival was 405 days versus 198 days for the HAI and systemic groups, respectively (P = 0.03).

The German Cooperative Group on Liver Metastases randomized 168 patients in a multicenter trial to one of three treatment groups: HAI of FUDR, HAI of 5-FU/LV, and systemic 5-FU/LV.[150] The median times to progression were 5.9, 9.2, and 6.6 months, and median survivals were 12.7, 18.7, and 17.6 months, respectively. Tumor response rates were 43.2%, 45%, and 19.7%, and development of extrahepatic disease was 40.5%, 12.5%, and 18.3%, respectively. It should be pointed out that only 70% of patients randomized to HAI of 5-FU/LV and 68.5% randomized to HAI of FUDR were actually treated.

The Medical Research Council/European Organization for the Research and Treatment of Cancer (MRC/EORTC) Colorectal Cancer Groups conducted a randomized trial of intravenous versus HAI administration of 5-FU/LV for 290 patients. Median and 2-year survival rates were 13.4 months and 23% in the intravenous arm as compared with 14.7 months and 20% in the HAI arm; however, 37% of patients allocated to the HAI arm did not receive the assigned treatment.[151]

The Cancer and Leukemia Group B (CALGB)[152] trial differs from the 135 other HAI studies in that it included the use of Dex in the HAI arm.[153] In patients, HAI FUDR + Dex + LV was compared to systemic-bolus 5-FU/LV. No crossover was allowed. The HAI group had a significant increase in survival of 24.4 months versus 20 months in the systemic group (P = 0.0034; Fig. 59-10 ). The time to hepatic progression was better in the HAI arm (9.8 months vs. 7.3 months in the systemic group; P = 0.034) but the time to extrahepatic progression was better in the systemic arm (14.8 months versus 7.7 months in the HAI group; P < 0.029). There was an improved quality of life in the HAI group, especially physical functioning, measured at 3 and 6 months (P = 0.024).

 
 

Figure 59-10  Kaplan-Meier overall survival by treatment arm for all patients entered.

 

 

There are differences between the CALGB study[152] and the European studies, which could explain differences in outcomes. The CALGB study used pumps instead of ports, and HAI therapy included FUDR with Dex to decrease toxicity.[153] Survival was based on intent to treat in all three studies, and the actual number of patients treated was much lower in the European studies, 66% in the German study and 63% in the English study, whereas it was 86% in the CALGB study. The CALGB study did demonstrate that regional therapy alone can improve survival over systemic 5-FU/LV with a survival similar to that which is seen utilizing the new agents. Randomized studies of HAI therapy versus the new therapies have not yet been done.

A meta-analysis combining the results of seven trials supports the use of HAI of FUDR in the treatment of nonresectable liver metastases from CRCs.[154] A significantly better local response rate of 41% was achieved with HAI of FUDR compared with a 14% response rate for systemic 5-FU. In addition, median survival time was increased in patients treated with HAI—16 months versus 13 months for those treated with 5-FU.

New Approaches to Decrease Hepatic Toxicity

The hepatic toxicity induced by HAI of FUDR could be related to portal triad inflammation, which could lead to ischemia of the bile ducts. Therefore, hepatic arterial administration of Dex might decrease biliary toxicity. In patients with established hepatobiliary toxicity from HAI, Dex promotes resolution of liver function abnormalities. In a randomized study of FUDR with Dex as compared to FUDR alone, there was a trend toward decreased bilirubin elevation in patients receiving FUDR + Dex compared with the group receiving FUDR alone (9% vs. 30%, respectively; P = 0.07).[153] Although the addition of Dex was not associated with a significant increase in the amount of FUDR that could be administered, the response rate was increased by 71% for the FUDR + Dex group as compared with 40% for the group receiving FUDR alone (P = 0.03). Survival was also improved: 23 months for patients receiving FUDR + Dex as compared with 15 months for those receiving FUDR alone (P = 0.06)

The use of circadian modification of HAI with FUDR is another method to decrease hepatic toxicity. In a nonrandomized study at the University of Minnesota, a comparison of constant (flat) infusion to circadian-modified hepatic arterial FUDR infusion was conducted.[155] The group with circadian modification received 68% of each daily dose between 3:00 pm and 9:00 pm. The patients with circadian modified infusion tolerated almost twice the daily dose of FUDR, with a decrease in hepatic toxicity compared with patients receiving flat infusions, but the study was not a prospective randomized study.

Another approach to decrease toxicity from HAI is to alternate drugs such as HAI FUDR with hepatic arterial 5-FU. A weekly hepatic arterial bolus of 5-FU does not cause hepatobiliary toxicity; however, it frequently produces treatment-limiting systemic toxicity or arteritis. Stagg and colleagues[156] used alternating HAI of FUDR (for 7 days) followed by a hepatic arterial bolus of 5-FU (weekly for 3 weeks every 35 days). The response rate was 51%, and median survival was 22.4 months. In contrast to the experience with single-agent HAI of FUDR, no patient on the alternating plan has had treatment terminated because of drug toxicity. Davidson and associates,[157] using a similar alternating regimen on 54 patients, produced a 54% response rate, with 3.5% of patients developing biliary sclerosis. Patt and coworkers,[158] using HAI of 5-FU and interferon in 5-FU- and LV-refractory patients, produced a 33% response rate with a median survival of 15 months and no hepatobiliary toxicity.

Warren and colleagues,[159] using weekly HAI 5-FU infusion and systemic LV, produced a 45% response rate in 31 patients without hepatic toxicity and mild systemic toxicity. Metzger and colleagues[160]treated 30 patients with an infusion of 5-FU + mitomycin C; the courses were repeated every 6 weeks. The median survival was 18 months, with a 57% partial response rate. No patients developed sclerosing cholangitis, but mucositis and leukopenia were seen. Catheter complications occurred in 33%, which led to premature termination of treatment in one-third of the patients. Schlag and associates,[161] using HAI of 5-FU via a surgically placed port connected to an external pump, reported a 27% partial response in 33 patients and a median survival of 14 months. Hepatobiliary toxicity was seen in only 5% of the patients. Table 59-15 lists some of the studies using 5-FU and other agents by this technique.


Table 59-15   -- Hepatic Arterial Therapy: 5-FU Plus Other Agents

Investigator

No. of Patients

Dose (mg/m2)

Response (%)

Median Survival

Cortesi et al[346]

109

5-FU 500 × 5

46

16

 

 

Cisplatin 24 × 5

 

 

Warren et al[159]

31

5-FU 1500 × 24 hr

48

17

 

 

LV 400 weekly × 6

 

 

Sugihara[347]

58

5-FU 360 × 7 then 100 × 7 days

50

11

Kerr et al[348]

43

LV 200

36

 

 

5-FU 400 then 5-FU 1800 × 22 hr

 

 

Howell et al[349]

40

LV 200

46

19

 

 

5-FU 400 then 5-FU 1600–22 hr every 2 wk

 

 

Patt et al[350]

48

IFN 5 mu

33

15

 

 

5-FU 1000 × 5 hr every wk

 

 

Borner et al[351]

28

5-FU 750 × 5 days

50

 

 

 

Methods to Increase Response Rate and Decrease Extrahepatic Disease

The potential benefit of multidrug hepatic arterial therapy has been evaluated. A randomized trial of a three-drug regimen (mitomycin C, bischloroethyl nitrosourea, and FUDR) compared to FUDR alone was conducted in 67 patients with disease progression after previous treatment with systemic chemotherapy.[162] The overall response rate and survival were similar. The response rates in both groups were higher than would be expected with a second systemic regimen after failure of first-line treatment. For patients who had been treated previously for metastatic disease, the response rate was 48% for the three drugs as compared with 24% for FUDR alone (P = 0.03).

HAI FUDR/Dex + mitomycin C[163] administered through the pump side port produced a 70% response rate in previously treated patients with a median survival of 19 months from the start of HAI therapy after progression on systemic FU/LV.

Sixty-four patients were treated with FUDR and LV by HAI. The overall response rate was 62%, but 15% of patients developed biliary sclerosis. The toxicity of the combination was higher than that of FUDR alone, but the survival was improved: 86% of the patients were alive at 1 year, 62% at 2 years, 33% at 3 years, and 10% at 5 years after treatment.[164]

To further increase response rate without an increase in toxicity, the combination of FUDR, Dex, and LV was tested.[165] The response rate was 78% in previously untreated patients, with only 3% of these developing sclerosing cholangitis. The median survival was 24.8 months, and the 1- and 2-year survivals were 91% and 57%, respectively. In patients who had received previous chemotherapy, the response rate was 52%.

The use of systemic therapy + HAI might produce a decrease in extrahepatic disease. In Safi and coworkers’ study[166] comparing FUDR (0.2 mg/kg/day for 14 of 28 days) with a combination of intraarterial FUDR (0.21 mg/kg/day) and intravenous FUDR (0.09 mg/kg/day) given concurrently for 14 of 28 days, the response rates were 60% for both arms of the study. The incidence of extrahepatic disease, however, was 56% for the group receiving intraarterial/intravenous treatment compared with 79% for those receiving HAI alone (P < 0.01). There was no difference in survival between the two groups (P = 0.08).[135] In Lorenz and colleagues,[150] study of 52 patients, combined HAI/intravenous did not increase survival or decrease the development of extrahepatic disease (60% and 62% for HAI/intravenous vs, HAI alone, respectively). Wanebo and colleagues[167] used continuous-infusion 5-FU and LV for 14 days alternating with intrahepatic FUDR, Dex, and LV. Two other studies produced median survivals of 12 and 18 months administering HAI FUDR + systemic 5-FU. [169] [170]

A phase I study of HAI FUDR combined with systemic irinotecan[170] in previously treated patients (45% had previous CPT-11) reported a response rate of 74%, a TTP of 8.1 months, and a median survival of 20 months. Thirteen of the sixteen patients with previous CPT-11 exposure responded to this regimen. Systemic Oxali + 5-FU/LV or Oxali + CPT-11 with concurrent HAI FUDR/Dex in 36 previously treated patients (74% had received prior CPT-11) produced response rates of 86% with a median survival of 36 months, and a 1-year survival of 80%.[171]

Toxicity of Hepatic Arterial FUDR Infusion

Most trials of HAI given via an internal implantable pump use FUDR or 5-FU. A summary of the toxicities encountered is listed in Table 59-16 . Myelosuppression, nausea, vomiting, and diarrhea do not occur with HAI of FUDR. If diarrhea does occur, shunting to the bowel should be suspected.[172] The most common problems with HAI FUDR are hepatic toxicity and ulcer disease ( Box 59-2 ).[173]Hepatobiliary toxicity is the most problematic toxicity seen with HAI FUDR. Most studies point to a combined ischemic and inflamma tory effect on the bile ducts as the etiology of this complication. The bile ducts derive their blood supply almost exclusively from the hepatic artery and thus are undoubtedly perfused with high doses of chemotherapy.[174]


Table 59-16   -- Hepatic Arterial FUDR Infusion with Internal Pump: Toxicities

Investigators

No. of Patients

Gastritis (%)

Ulcer (%)

SGOT (%)

Bilirubin (%)

Diarrhea (%)

Biliary Sclerosis (%)

Niederhuber et al[333]

70

56

8

32

24

Balch and Urist[334]

50

6

23

23

0

Kemeny et al[335]

41

29

29

71

22

0

5

Shepard et al[336]

53

20

49

24

Cohen et al[337]

50

40

10

25

Weiss et al[338]

17

50

11

80

23

23

Schwartz et al[339]

23

53

77

20

10

Johnson et al[340]

40

8

50

13

-0

5

Kemeny et al[175]

31

17

6

47

-8

29

Hohn et al[342]

61

35

2

0

78

11

29

SGOT, serum glutamic-oxaloacetic transaminase.

 

 

 

Box 59-2 

DOSE MODIFICATION FOR HEPATIC TOXICITY

SGOT

Bilirubin

Alkaline Phosphatase

FUDR Dose Reduction (%)

<2 × baseline

80

3 × baseline

or 1.5×

or 1.5×

50

>3 × baseline

or 2×

or 2×

Hold

 

 

Baseline is day 1 or previous FUDR dose. If elevations continue on reduced doses, decrease duration of FUDR to 1 week. If bilirubin is >2 mg/dL, do not repeat treatment for 2 months. If bilirubin levels return to and remain normal, use 25% of the previous FUDR dose for 1 week.

Clinically, biliary toxicity manifests as elevations of serum glutamic-oxaloacetic transaminase (SGOT), alkaline phosphatase, and bilirubin. Elevation of SGOT is an early manifestation of toxicity; elevation of alkaline phosphatase or bilirubin is evidence of more severe damage. In the early stages of toxicity, hepatic enzyme elevation returns to normal when the drug is withdrawn and the patient is given a rest period, whereas in more advanced cases it does not resolve.

In patients who develop jaundice, endoscopic retrograde cholangiopancreatography (ERCP) might demonstrate lesions resembling idiopathic sclerosing cholangitis in 5% to 29%.[175] Because the ducts are sclerotic and nondilated, sonograms usually do not show dilation. In some patients, the strictures are more focal and usually worse at the hepatic duct bifurcation; drainage procedures either by ERCP or by transhepatic cholangiogram could be helpful.[176] Duct obstruction from metastases should first be excluded by CT of the liver.

Close monitoring of liver function tests is necessary to avoid the biliary complications. If serum bilirubin rises to 3 mg/dL or higher no further treatment should be given until the bilirubin returns to normal, and then only after a long rest period to prevent the development of sclerosing cholangitis. With proper monitoring, this complication occurs in less than 10% of patients.

Severe ulcer disease results from inadvertent perfusion of the stomach and duodenum with drug via small collateral branches from the hepatic artery and can be prevented via careful dissection of these collaterals at the time of pump placement. Even without radiologically visible perfusion of the stomach, however, mild gastritis and duodenitis still can occur. HAI 5-FU causes less biliary problems and more arteritis ( Box 59-3 ).[156]

Box 59-3 

INTRAHEPATIC THERAPY MANAGEMENT OF TOXICITY

Ulcer: Denude vessels that supply stomach and duodenum

Hepatitis: Monitor SGOT, alkaline phosphatase; decrease or hold dose for enzyme elevations.

Biliary sclerosis: Monitor liver enzymes, document by ERCP; if present, do not repeat treatment even if enzymes normalize.

Cholecystitis: Remove gallbladder.

Diarrhea: May be misperfusion or very extensive liver disease (repeat hepatic flow scan).

Complications of Hepatic Artery Infusion Pump Placement and Use

Incomplete or extrahepatic perfusion is usually a surgical problem. It is necessary to mobilize the entire gastroduodenal artery and ligate all branches supplying the stomach, duodenum, bile duct, or pancreas. The postoperative 99mTc macroaggregated albumin perfusion scan, when compared with the liver scan, can show incomplete perfusion of the liver from either an accessory or replaced hepatic artery not detected preoperatively or an incompletely isolated hepatic artery that is supplying vessels to the pancreas or gastrointestinal tract. These abnormalities can be assessed by performing a side port pump injection ( Box 59-4 ). (Most implantable pumps have a main chamber with a central port in which the drugs, such as FUDR, are placed for constant infusion. These pumps also have a side port that bypasses the main chamber and goes directly into the main catheter.) Often an accessory left hepatic artery arising from the left gastric artery, or an accessory right hepatic artery arising from the superior mesenteric artery, has not been identified. In such instances it might be possible to embolize the accessory vessel, which will allow collateral flow to develop from the main hepatic artery. A common technical error during catheter insertion is placing the catheter too far from the junction of the gastroduodenal artery and hepatic artery. The segment of gastroduodenal artery that is continuously exposed to full-dose chemotherapy will eventually thrombose. There is a learning curve for both the surgeons and the medical oncologists. In one series, technical complications were seen in 37% of patients with inexperienced surgeons and in 6.6% of patients with experienced surgeons.[177]

Box 59-4 

TECHNICAL ASPECTS OF PUMP INSERTION

  

1.   

Rule out extrahepatic disease by colonoscopy, chest radiograph, or CT scan of abdomen and pelvis.

  

2.   

Intraoperatively, biopsy samples should be taken from portal lymph nodes to rule out extrahepatic disease.

  

3.   

Celiac and superior mesenteric artery arteriography should be done to identify arterial anatomy of liver and vessels to stomach and duodenum and pancreas.

  

4.   

Portal vein must be patent.

  

5.   

Place catheter into gastroduodenal artery, not directly into hepatic artery, which could lead to thrombosis; secure catheter with nonabsorbable ties.

  

6.   

Identify and ligate branches to pancreas and duodenum.

  

7.   

Check whole-liver perfusion with 5 mL of fluorescein through side port.

  

8.   

Postsurgery macroaggregated albumin scan should be performed through side port of pump to check perfusion of the liver and to ensure no extrahepatic perfusion.

Hepatic Arterial Infusion: Conclusions

HAI has several advantages. From a pharmacologic standpoint, HAI is more effective than systemic therapy, in that higher drug levels are achieved at the sites of metastatic disease. Utilizing agents with high hepatic extraction virtually eliminates the systemic toxicity observed with “standard” chemotherapy.

Should every patient with unresectable liver metastases and no extrahepatic disease undergo HAI? Hepatic arterial therapy produces higher response rates compared with systemic therapy. Older randomized studies with crossover design did not prove a survival advantage. In ten randomized trials, the response rate was higher using HAI compared with systemic therapy (see Table 59-14 ). The time to hepatic progression was significantly longer in the HAI groups as compared with the systemic groups ( Table 59-17 ). The randomized pump studies do not evaluate the issue of survival clearly, for the following reasons: (1) In some studies a crossover was allowed; (2) the number of patients was small; (3) positive portal nodes were included in the HAI-treated groups. Recent HAI trials demonstrate median survivals of 24 to 36 months. Therefore, in a patient with extensive hepatic disease (30% to 60% liver involvement) who can tolerate an operation, hepatic arterial therapy combined with systemic chemotherapy or systemic chemotherapy alone can be used as the first treatment. Patients who start with systemic therapy and whose tumor fails to respond can then still undergo HAI therapy ( Fig. 59-11 ).Table 59-18 lists trials using HAI therapy after progression on systemic therapy. In institutions in which physicians are not familiar with placement of the hepatic arterial catheter and the pump or with the management of patients on regional infusion, systemic chemotherapy is a reasonable option.


Table 59-17   -- Randomized Studies of HAI versus SYS: Time to Hepatic Progression

 

 

(MO)

 

Group

No. of Eligible Patients

HAI

SYS

P Value

MSKCC[341]

99

9

6

 

NCOG[342]

117

13

6.7

 

Mayo Clinic[344]

64

15.7

6

 

France[345]

163

15

6

 

CALGB[152]

135

9.8

7.3

.034

HAI, hepatic arterial infusion; SYS, systemic chemotherapy.

 

 

 

 
 

Figure 59-11  A, Patient's CT before treatment, demonstrating massive liver involvement. B, Patient's CT after hepatic arterial therapy, showing an excellent partial response after HAI FUDR and LV.

 

 


Table 59-18   -- Hepatic Arterial Infusion in Previously Treated Patients: Second-Line Therapy

Regimen

No. of Patients

Partial Response (%)

1-Year Survival (%)

HAI alone

 

 

 

 FUDR alone[352]

49

67

33

 FUDR + Mit C+ BCNU[352]

45

70

47

 FUDR + LV + Dex[166]

29

66

52

 Mit + FUDR + Dex[164]

37

77

65

 CPT-11[353]

10

33

 5-FU/LV + Cisplatin[354]

28

48

58

HAI + SYS

 

 

 

 FUDR + Dex + SYS–CPT-11[171]

56

74

84

 FUDR + Dex + SYS–Oxali[171]

36

80

86

 Oxali + SYS–FU/LV[355]

22

68

 Oxali + SYS–FU/LV[356]

28

64

85

FUDR alone[340]

 

 

 

BCNU, bischloroethyl nitrosourea or carmustine; CPT-11, irinotecan; Dex, dexamethasone; 5-FU, 5-fluorouracil; FUDR, 5-fluoro-2-deoxyuridine; HAI, hepatic arterial infusion; LV, leucovorin; Mit C, mitomycin C; Oxali, oxaliplatin; SYS, systemic chemotherapy.

 

 

 

Because the liver is often the initial site of metastatic disease in patients with CRC, early intensive therapy with surgical resection or intrahepatic infusion (or both) at a time when the tumor burden is small might prevent the progression of metastases to other sites. Although hepatic arterial therapy is applicable to only a minority of patients with metastatic CRC (those with only hepatic metastases), it might be the best available therapy for these patients.

In the future, molecular markers could help determine the optimal type of treatment. In one small study, patients with low thymidylate synthetase levels had a fourfold greater chance of responding after 5-FU-based infusion.[178] In another study, patients with mutated p53 had a shorter survival.[179]

The use of HAI for other tumor types has not been studied as extensively. Some of the breast studies using HAI are shown in Table 59-19 . Most of these studies are on patients whose tumors are failing to respond to systemic therapy, and yet response rates of 50% or greater are seen in five of the eight studies.


Table 59-19   -- Hepatic Arterial Infusion for Breast Cancer,[*]

Investigators

No. of Patients

Drugs

Partial Response (%)

Survival (mo)

Fraschini et al[357]

34

Cisplatin + vinblastine

33

11

Estape et al[358]

16

VP-16–cytoxan

50

16

Fraschini et al[359]

25

Vinblastine

52

11[†]

Fraschini et al[360]

26

Cisplatin

19

11

Maral et al[361]

15

Mit C + FUDR

53

18[‡]

Arai et al[362]

56

FU + doxo + Mit C

81

12.5

Tada et al[363]

45

Doxo + Mit C

37

7.5

FU, 5-fluorouracil; FUDR, fluorodeoxyuridine; Mit C, mitomycin C.

 

*

Most studies consist of patients whose tumor progressed after previous systemic chemotherapy.

Mean survival only for responders.

Some patients in study also had colon cancer.

 

Trials have been done using HAI therapy for cholangiocarcinoma and HCC (Tables 59-20 and 59-21 [20] [21]). At MSKCC, HAI therapy alone produced a partial response of 47% in cholangiocarcinoma and 28% in HCC.[180] Yoshikawa and associates, using HAI, epirubicin, pararubicin, and carboplatin, had an objective response of 35.4% and 2-year survival of 30% in 319 patients.[181]


Table 59-20   -- Biliary Tract Carcinoma—Hepatic Arterial Infusion Trials

Study

No. of Patients

Therapy

Response Rate (%)

Median Survival

Warren[364]

15

HAI Doxorubicin

60

NR

Makela and Kairaluoma[365]

27

HAI Mit C

48

14 mo

Melichar et al[366]

32

HAI 5-FU/FA ± Cisplatin

NR

14 ± 17 mo

Cantore Update[367]

26

HAI Epirubicin/Cisplatin

35

NR

 

 

Systemic 5-FU IVCI

 

 

Danso[180]

23

HAI FUDR

47

NR

 

 


Table 59-21   -- Hepatic Artery Infusion Chemotherapy Trials in Hepatocellular Carcinoma

Study

No. of Patients

Treatment

PR (%)

Median OS

Okudaet al[368]

31

HAI cisplatin, 5-FU

29

NR

Shildt and Baker[369]

30

HAI FUDR, Dox, Strepto

10

NR

Fazio et al[370]

30

HAI Mit C, Cisplatin, 5-FU

11

8.6 mo

Patt et al[158]

29

HAI FUDR, LV, Dox, cisplatin

41

15 mo

Wellwood et al[371]

28

HAI FUDR

54

NR

Carr and Dvorchik[372]

26

HAI cisplatin

42

NR

Cheng et al[373]

16

HAI cisplatin

19

NR

Urabe et al[374]

15

HAI MTX, 5-FU, cisplatin, IFN

47

NR

Atiq et al[375]

10

HAI FUDR, Mit C, IFN

50

14.5 mo

Dox, doxorubicin; 5-FU, 5-fluorouracil; FUDR, 5-fluoro-2-deoxyuridine; HAI, hepatic artery infusion; IFN, interferon; Mit C, mitomycin C; LV, leucovorin; MTX, methotrexate; Strepto, streptozotocin.

 

 

 

Neoadjuvant Chemotherapy

Neoadjuvant chemotherapy has been evaluated as a way to decrease hepatic tumor burden for the purpose of proceeding with hepatic resection. Although no randomized prospective trial has been completed testing this as a treatment strategy, the results of a retrospective analysis have been presented, and the results seem promising.

Giacchetti and colleagues[182] conducted a retrospective review of 151 patients with unresectable liver-only metastases from CRC who were treated with 5-FU/LV and Oxali followed by attempted liver resection. The criteria used to define unresectability were as follows: (1) more than four liver metastases (30%), (2) single tumor larger than 5 cm (34%), (3) tumor in both hepatic lobes, (4) invasion of the intrahepatic vascular structures, and (5) high percentage of liver involvement (>25% liver involvement; 48%). Their analyses revealed that of this group, 77 patients (51%) became resectable after neoadjuvant therapy. The 5-year survival rate for the 77 patients who underwent hepatic resection was 50%, progression-free survival was 17 months, and 72% had relapsed within a median of 12 months.[183] [184] Adam and coworkers,[84] from the same group, reviewed the records of 1104 patients with unresectable disease who received mainly FOLFOX as neoadjuvant therapy; 12.5% became resectable and the 5-year survival of these patients was 34%, similar to what could be obtained with patients who were initially resectable.

Several trials were designed to address resectability rate after chemotherapy that varied with patient selection. The Mayo Clinic trial considered patients to be unresectable if they had (1) involvement of all three major hepatic veins, portal vein bifurcation, or the hepatic vena cava; (2) involvement of the main right or main left portal vein and the main hepatic vein of the opposite lobe; (3) disease requiring more than a right or left trisegmentectomy; or (4) six or more metastatic lesions distributed diffusely in both lobes of the liver. In patients treated with preoperative Oxali/5-FU/LV, 38% of 44 became suitable for resection, but after resection, 73% recurred in the liver. Median survival was 26 months.[184] A retrospective review of these data demonstrated that 10% of patients were actually resectable before neoadjuvant therapy.

Pozzo and colleagues treated 40 patients with unresectable disease, meaning: (1) six metastases or three per lobe, (2) size larger than 5 cm for one lesion if six metastases were present, (3) contiguity with two hepatic veins, inferior vena cava or liver hilum with neoadjuvant FOLFIRI.[185] Thirty-two percent underwent liver resection with a DFS of 28 months. Other trials evaluating neoadjuvant chemotherapy are in Table 59-22 .


Table 59-22   -- Prospective Trials Evaluating Neoadjuvant Systemic Chemotherapy for Unresectable Disease: Respectability Rate

Investigator

Treatment

No. of Patients

Resectability Rate (%)

Wein[376]

Oxali + FU/LV

53

11

Pozzo[185]

FOLFIRI

40

27.5

Alberts[184]

FOLFOX

42

33

Gaspar[377]

FOLFOX

37

27

Quenet[378]

Oxali/CPT-11/FU/LV

34

37.5

Abad[379]

Oxali/CPT-11/FU

47

30

Falcone[380]

Oxali/CPT-11/FU

39

30

Falcone[380]

FOLFIRI

42

12

Ho[381]

FOLFIRI

40

10

Masi[382]

FU/LV/Oxali + CPT-11

74

26

Wein[383]

FU/LV

53

11

Rougier[384]

Cetuximab + CPT-11/FU/LV

23

30

 

 

Rougier and associates treated 23 patients with cetuximab and FOLFIRI, and seven (30%) became resectable.[186] Oxali/Irinotecan (CPT-11) + HAI in 44 clearly unresectable patients enabled 35% to become resectable (56% even though the majority had received this after second- or third-line therapy). The median survival of the entire group is 36 months, and median survival for the resected group has not been reached. In a Japanese study, 51 patients with unresectable disease were treated with HAI/5-FU and systemic UFT. The 3- and 5-year survivals were 58% and 42% for the resected group. In Italy, 244 patients were randomized to FOLFIRI or FOLFIRI + Oxali (Folfoxiri). In patients with only liver metastases, the resection rate was 12% and 36% for the FOLFIRI and Folfoxiri regimens, respectively.[187]

In making decisions about the timing of liver resection and pre- or postoperative chemotherapy, there are several considerations. The advantages and disadvantages are listed in Table 59-23 . For patientswith borderline resectability, chemotherapy can be done first but only for short periods of time, and as soon as response is noted, surgery should be performed. Adam reported prolonged chemotherapy (>12 cycles) was related to longer hospital stays after liver resection.[188] For clearly unresectable disease, patients should enter protocols that are addressing the question of how best to decrease liver metastases to allow resection. In patients with synchronous CRC, there are those who advocate for simultaneous resection of both the primary and the metastatic disease, those who recommend chemotherapy first, [190] [191] and those who state both surgeries can be done simultaneously.


Table 59-23   -- Pros and Cons of Preoperative Chemotherapy

  

 

PROS

  

   

Decrease size

  

   

Control micrometastatic disease

  

   

Assessment of chemotherapy activity

  

   

Better chemotherapy tolerance

  

   

Surrogate marker for success of liver toxicity

  

 

CONS

  

   

Liver toxicity

  

   

Steatosis

  

   

Sinusoidal dilatation

  

   

Risk of progression or growth of new sites

  

   

Clinical response may make surgery more difficult

  

   

Secondary splenomegaly

  

   

Selection of resistant clones

 

 

Supporting the concept of controlling micrometastatic disease, the Liver Met Survey group found patients with more than five metastases survived longer if they were given neoadjuvant chemotherapy, with a 5-year survival of 22% versus 12% (P = 0.07) for the preoperatively treated and preoperatively nontreated groups, respectively. In another retrospective review at MSKCC, preoperative chemotherapy did not improve survival. In a review of 230 patients who received HAI + systemic therapy after liver resection,[191] the median survivals for patients who did or did not receive preoperative (neoadjuvant) therapy were 63 and 115 months, respectively(P = 0.26).

If metastases that disappear on CT remain viable, preoperative chemotherapy may be a disservice. In a review by Benoist and coworkers on 586 treated patients,[192] 38 patients had disappearance of at least one lesion on CT. Pathologic examinations of sites with a complete response showed that in 11 of 15 (80%) there were viable tumor cells. Areas that could not be found were closely followed, and in 23 of 41 sites (47%), tumor reoccurred.

As use of preoperative chemotherapy increases, reports are emerging about liver toxicity from preoperative chemotherapy. There is an increased incidence of steatosis, sinusoidal abnormalities, venoocclusive disease, and steatohepatitis. In the Rubbia-Brandt and colleagues[193] report on 153 patients undergoing liver resection, 51% of the 87 patients who have received chemotherapy before resection had sinusoidal dilation and 78% who had prior Oxali showed striking sinusoidal alterations, with venoocclusive fibrosis in 48%. Kooby and associates reported on 325 patients with fatty livers undergoing resection at MSKCC and found that those treated with preoperative chemotherapy were more likely to have steatosis (66%), with marked steatosis being an independent predictor of complications following hepatic resection.[194] Vauthey and coworkers reported a 20% incidence of steatohepatitis in patients receiving preoperative CPT-11 versus 4.4% in those who received no preoperative chemotherapy, and an 18.9% incidence of sinusoidal dilation in those receiving preoperative Oxali versus 1.9% in those receiving no preoperative chemotherapy. Patients with steatohepatitis had an increased 90-day mortality, 14.7% versus 1.6% (P = 0.001).[195]

Is neoadjuvant chemotherapy useful for resectable liver metastases? The EORTC study is looking at preoperative and postoperative FOLFOX versus observation. They have randomized 364 patients to (1) six cycles of FOLFOX before and after surgery or (2) surgery alone.[196] Presently, in the group receiving preoperative chemotherapy, 7.7% progressed before surgery and 11% were not able to undergo resection. It is too early to look at the DFS and survival. For clearly resectable disease, until we know more from these studies, surgery should be done first, followed by adjuvant therapy. Figure 59-12provides a treatment algorithm for liver metastases.

 
 

Figure 59-12  Treatment of liver metastases.

 

 

HEPATIC ARTERIAL EMBOLIZATION

Because both hepatic metastases and primary liver tumors derive their blood supplies from the hepatic artery, hepatic arterial ligation (HAL) and embolization have been used to reduce the tumor's blood supply.[197] In most hepatic tumors, HAL produces only a transient benefit because of the rapid development of collateral vessels. In vascular neuroendocrine tumors, HAL produces objective tumor reduction, whereas in colon metastases (which are typically less vascular), minimal regression occurs.

The development of a collateral blood supply might be minimized by injecting vasoocclusive particles into the hepatic artery (hepatic artery embolization, HAE), which also provides the opportunity for retreatment, because only the microvasculature is occluded.[103] In one study of 61 individuals with liver metastases from colorectal carcinoma, patients were randomized to HAE, HAL + microspheres (HAE), or no further treatment. The median survivals were 8.7, 13, and 9.6 months, respectively, suggesting that embolization or ligation alone had no effect on survival for patients with CRC.[198] Gerard and colleagues[199] randomized 67 patients to HAL alone versus HAL and portal vein infusion of 5-FU. The median survival in both groups was 12 months; among patients whose median extent of liver involvement was 30%, only one patient responded. Both studies suggest that this technique is not useful for CRC.

HAE has a definite role in highly vascular tumors, such as neuroendocrine tumors of the liver.[200] These tumors usually grow slowly, and a reduction in tumor bulk can result in significant palliation. Ajani and associates[201] treated 22 patients who had islet cell carcinoma with HAE using polyvinyl alcohol particles (Ivalon) and gelatin sponge particles (Gelfoam). The median survival in this study was 33 months from initiation of embolization (range, 1 month to 72 months); 12 patients had a partial response associated with subjective improvement and decrease in hormone levels. Other studies using HAE are listed in Table 59-24 ; all demonstrate an improvement in symptoms and a decrease in hormone levels. Gelfoam (size 1–2 mm), one of the agents used for embolization, does not lead to peripheral vascular occlusion and has an inconsistent duration of occlusion (due to absorption), whereas Ivalon (size 150–500 μm), a smaller particle, allows for more peripheral occlusion and is not absorbable, allowing for a more persistent arterial occlusion. [202] [203] [204] A collagen particle (Angiostat; size 20–250 μm) and a biodegradable albumin microsphere (Spherex; size 15–40 mm) are currently being tested, especially in combination with chemotherapy.


Table 59-24   -- Neuroendocrine Tumors

Investigators

Agent

No. of Patients

Biochemical or Symptomatic Response (%)

Tumor Response (%)

Hepatic artery ligation

 

 

 

 

 Martin et al[206]

 

8

76

 

 Moertel[207]

 

10

70

 

 Melia et al[208]

 

6

50

17

Hepatic artery embolization

 

 

 

 

 Carrasco et al[202]

Gelfoam

23

87

 Ajani et al[201]

Ivalon + Gelfoam

22

60

 

 Marlink et al[203]

Gelfoam

10

100

90

 Maton et al[385]

Lyodura

13

76

 

 

Interference with hepatic blood flow can exacerbate the underlying liver disease and can be dangerous in patients with portal venous thrombosis, which is present more often in patients with primary liver tumors than in those with metastatic disease. In a study by Carr and coworkers,[204] four patients with HCC had reversal of tumor-induced portal vein thrombus. The researchers obtained objective responses in 22 of 35 patients (63%) using Spherex (biodegradable) starch microspheres, doxorubicin, and cisplatin. There is clear evidence that arterial embolization causes antitumor effects, but some randomized studies have not shown increased survival.[205]

Complications of Embolization

The complications of embolization are nausea, vomiting, fever, pain, and changes in liver function tests ( Table 59-25 ). Problems less commonly encountered include the following: injury of the gallbladder by retrograde flow through the cystic artery, ischemic necrosis of the bowel by embolization of one of the vessels to the intestinal tract, pancreatic infarction and pancreatitis by embolization of one of the pancreatic vessels, and dyspnea by embolization of the lungs. [202] [203] [204] [207] [208] [209]


Table 59-25   -- Complications and Management of Hepatic Arterial Embolization

COMPLICATION

Pain

Fever

Nausea and vomiting

↑WBC ↑LDH ↑SGOT

Cholecystitis

Hepatic gas formation and abscess

Renal insufficiency

Ileus

PRE-EMBOLIZATION

Hydration

Allopurinol

Somatostatin (only in neuroendocrine carcinomas)[*]

Analgesics just before procedure

POSTEMBOLIZATION

Analgesics

Follow WBC, LDH, SGOT, creatinine

Treat by appropriate measures nausea, fever, abscess, infection, ileus

LDH, lactate dehydrogenase; SGOT, serum glutamic-oxaloacetic transaminase; WBC, white blood cell count.

 

*

150–250 mg SC, every 6–8 hrs before procedure.

 

In neuroendocrine tumors responding to HAE, rapid cell death can result in tumor lysis syndrome with symptomatic hyperuricemia, leading to uric acid nephropathy and oliguria. Vigorous hydration and prophylactic allopurinol might prevent this problem. In carcinoid tumors, HAE could cause a life-threatening carcinoid crisis from the rapid release of hormones from tumor cells. Somatostatin analogs may be given, either before the procedure or if a carcinoid crisis should occur.[209] To avoid some of the serious complications of HAE, patients with cirrhosis, portal vein occlusion, and biliary tract obstruction are usually excluded.

CHEMOEMBOLIZATION

An extension of the work with embolization is chemoembolization. This process involves a local entrapment of drug in the embolization agent and provides a prolonged exposure of the tumor to the drug locally with less systemic drug circulation. A nonrandomized study by Daniels and colleagues[210] suggested that the addition of chemotherapy to the embolic agent (angiostat) produced an increase in response rate over the embolic agent alone. In a study by Venook and associates,[211] 51 patients with unresectable HCC were treated with Gelfoam and a mixture of three drugs—doxorubicin, mitomycin C, and cisplatin—given via a percutaneous hepatic artery catheter. Twelve (24%) had a partial response, and tumor liquefaction was noted in 70% of patients on CT, with a more than 50% reduction in α-fetoprotein in 68% of patients. Using this technique and the same drugs, these investigators also treated liver metastases from neuroendocrine tumors. In 12 patients with a median liver involvement of 60%, 33% had a partial response with a reduction in hormone levels.[160] Median survival from treatment initiation was 7 months (range 3 months to 3 years). Lipiodol has been found to remain selectively in the primary and secondary liver cancers when injected into the hepatic artery, allowing visualization of tumors as small as 4 mm.[161] Thus, Lipiodol can be used to deliver either chemotherapy or local radiation by combining with an agent such as iodine-131 (131I).[212]

Chemoembolization is rarely used for CRC, because median survivals are usually not increased; average survival is approximately 9 months.[213] Adding systemic therapy to hepatic chemoembolization could improve results. In one trial using regional cisplatin in a polyvinyl alcohol suspension with systemic 5-FU, the partial response rate was 40%, and median survival was 19.3 months.[214]

Another form of chemoembolization is to enclose the chemotherapeutic agents in a microsphere.[215] Degradable starch microspheres injected intra-arterially are trapped in an extracapillary network formed in liver metastases. [216] [217] Drug dissolved in the microsphere suspension is retained in the blood vessels of the target organ as long as the blood flow is blocked and then gradually releases the chemotherapeutic agents, resulting in a longer duration of tumor exposure to the drug. The most appropriate agents for microspheres would be those that, like mitomycin C, are preferentially toxic to cells under hypoxic conditions. Another useful drug to use with microspheres is doxorubicin. In a rabbit study, the mean tumor drug level was significantly higher when this drug was used with the microsphere compared with doxorubicin alone, whereas hepatic uptake of the drug by normal tissue was similar in the two groups.[217] Monoclonal antibodies can also be attached to the microspheres.[218]

Radioembolization attempts have been made using glass microspheres containing 90Y, a β-emitter with tissue penetrance of 2.5 mm. Andrews and coworkers[218] reported 5 of 23 responses with yttrium embolization. A study of 131I-labeled lipiodol administered to 20 patients (15 HCC, 5 metastatic) produced an α-fetoprotein drop in 11 of 12 patients and a response in 9.[219] The response rates for radioembolization of metastatic tumors are far less than those reported for treatment of primary HCC, perhaps because most metastatic tumors except for neuroendocrine primaries are far less vascular than HCC.

Arterial embolization is commonly used for HCC. Responses and increase in time to tumor progression have been documented. [206] [221] [222] [223] Meta-analysis showed an increase in survival with chemoembolization compared with conservative management.[223]

CRYOSURGERY

Cryosurgery is an in situ destruction of tissue using subzero temperatures. The rapid freeze/thaw of tissues results in cellular damage and death. One advantage of cryosurgery is the ability to use local treatment without sacrificing normal tissue. Among the difficulties with this technique are defining the full extent of the tumor and the inability to monitor the amount of freezing, and thus the possibility of overtreatment of surrounding vulnerable normal tissue. Two technical developments have improved the use of cryosurgery:

  

1.   

Cryoprobes cooled by liquid nitrogen allow more precise freezing, even within the liver.

  

2.   

Intraoperative ultrasound allow precise placement of the cryoprobe and more accurate monitoring of the freezing process.

Cryosurgery has been used intraoperatively but also can be used percutaneously. In a series of 32 patients with liver tumors (24 with CRC), 28% remained free of disease for 5 to 60 months.[224] In another series using intraoperative cryosurgery on 18 patients who had metastatic CRC with 1 to 12 lesions, Onik and coworkers[225] reported that 4 patients had complete remission with a median survival of 28 months, whereas 14 patients were considered inadequately treated and had a 21-month median survival.

Weaver and colleagues[226] treated 47 patients with cryosurgery with occasional operative resection. The number of metastases ranged from 1 to 12. The 2-year survival was 62%. Morris and Ross,[227]reporting on 67 patients, noted that 75% of patients undergoing cryosurgery had an increase in CEA by 6 months later. Occasionally, the surgeon feels that all disease has been destroyed but, as seen by this study's PET results after cryosurgery, tumor can still be present even though the surgeon feels that no disease has been left behind.

The 2-year survival after cryosurgery varies from 72% to 12%. The Boston series reported the highest survival, which might reflect the type of patient being selected for cryosurgery (i.e, lesser extent of disease and a smaller number of metastases).[228] Adam and associates[229] reported a 2-year survival rate of 50% for patients with colorectal metastases as compared with 67% for patients with HCC. They reported a local recurrence rate of 44% for the colorectal patients. Because local recurrence is high, the use of HAI after cryosurgery could be useful. One small, nonrandomized study doubled survival with the use of HAI after cryosurgery.[230] Other trials are now evaluating the use of HAI with or without systemic therapy. In a series of 185 nonrandomized patients, 71 received adjuvant CPT-11 and/or HAI of FUDR after cryosurgery. Two-year survival was 75% for patients receiving postcryosurgery therapy as compared with 35% if no adjuvant therapy was given.[231] Cryoablation can also be used after hepatic resection with close margins or to remove central lesions. Cryoprobes can be used as a handle to assist in segmental resections. An ice ball is produced with 1-cm margins around the tumor; then the probe is used for traction so that a segmental resection can be performed.[232]

The addition of cryosurgery to conventional surgical procedures was evaluated by Seifert and coworkers[118] in a randomized study. Those receiving surgical procedures plus cryosurgery had similar survival but an increase in liver recurrence for the cryo group.

Cryosurgery does involve some technical issues, including the following:

  

   

Adequate hydration before surgery, because myoglobinuria and tumor lysis can occur.

  

   

Attention to bile ducts, because biliary fistula can occur.

  

   

Two freeze/thaw cycles are preferred.

  

   

The probe should not be pulled or twisted vigorously, because that could cause cracking.[233]

Complications include hepatic cracking secondary to the thermal stresses that occur during rapid freezing; these are usually associated with hemorrhage, which could require packing. Other complications include biliary fistula requiring percutaneous drainage (which occurred in one patient) and myoglobinuria, resulting in acute tubular necrosis.

Published data do not support the use of cryosurgery in patients with resectable disease outside of a clinical trial.[234]

RADIOFREQUENCY ABLATION AND MICROWAVE COAGULATION

Just as tumors can be destroyed by cold, they can also be destroyed by heat. Techniques such as RFA and microwave coagulation (MC) have been used to destroy tumors. RFA involves placing within the tumor a small electrode, which is used to deliver energy to the tissue. The radiofrequency current generates ionic agitation, which is converted into frictional heat and results in breakdown of proteins and cellular membranes. The larger tumors can be destroyed by cryoablation. For RFA, tumors must be less than 4 or 5 cm in size. During the ablation, a hyperechoic area is formed around the tip of the needle, which corresponds to the area treated. It is sometimes difficult to evaluate whether all tumors have been treated. One of the advantages of RFA as opposed to cryoablation is that it can be performed percutaneously, because the probes are 10 mm in length. Solbiati and colleagues[235] treated 109 patients with colorectal metastases. He found a local control of 70%. Recurrence was significantly more frequent among patients with lesions larger than 3 cm. New metastases developed in 50% of patients, and survival rates were 67% and 33% at 2 and 3 years, respectively.[235] Bilchik and associates[236]proposed an algorithm for unresectable hepatic neoplasms, using cryosurgery for larger lesions and radiofrequency for tumors smaller than 3 cm, because local recurrences occurred in 38% of those receiving RFA and only in 17% of patients receiving cryosurgery. Among the most useful situations in which to use RFA is for patients with HCC who also have cirrhosis. Curley and coworkers[237]presented a series of 110 patients with cirrhosis who received RFA for HCC, with no recurrences occurring in 50%.

Cancer cells could be more sensitive than normal cells to heat due to the decreased vasodilation capacity of the neurovascular bed.[238] MC was initially developed for coagulation. When microwaves are applied to living tissue, they act mainly on the watery component. Using a probe to deliver 80-watt output for a 30-sec duration creates a column of coagulated area of 10 mm. In 19 patients with HCC, 28 of 31 nodules underwent complete tumor ablation. Ten of the 19 patients are still free of disease (followup, 14–64 months). Advocates of this therapy suggest that MC does not have inhomogeneous distribution within the tumor as seen with percutaneous ethanol injection therapy (PEIT).[239] MC is useful only in very small tumors (<3 cm). Both modalities (MC and PEIT) might be more useful if combined with embolization. Lesions near hilar structures are not good candidates for MC, but lesions adjacent to hepatic veins can be treated.

A Italian study on 423 patients treated with RFA reported 3- and 5-year survivals of 47% and 24%, respectively,[240] which is similar to surgical series. In a series of 25 patients where resection was not performed because the lesion was close to a major vessel, the 3-year survival with RFA alone was 52%.[241] Elias and colleagues used RFA instead of repeat resections for the treatment of recurrence after liver resection in 47 patients. They found that this increased the percentage of local cures and decreased the need for resection.[242] Livraghi and associates evaluated the role of RFA in 88 patients who were waiting for resection. Fifty-three patients had complete tumor ablation by RFA, and 16 (30%) remained tumor-free. Among these 53 patients with complete ablation, 98% were spared surgical resection—44% because they remained tumor-free and 56% because they developed additional metastases. Lesions in 35 patients (40%) demonstrated local tumor recurrence, and 15 developed unresectable disease.[243]

In the Cleveland Clinic 135 patients received RFA because they were not good candidates for surgical resection, and the median overall survival was 28.9 months. Predictors of survival included size of the lesion and baseline CEA values. There was a median survival of 38, 34, and 21 months for lesions smaller than 3 cm, 3 to 5 cm, and larger than 5 cm, respectively (P = 0.03). Median survivals of 34 and 16 months were seen with CEAs of less than 200 ng/mL versus greater than 200 ng/mL, respectively (P = 0.01). In the Cox proportional hazards model, only the size of the largest lesion (>5 cm) was found to be a significant predictor of survival.[244] Other studies have reported size as an important predictor of recurrence.[245]

At M.D. Anderson Cancer Center 348 patients with liver metastases from CRC were treated for cure: 190 had resection only, 101 had RFA and resection, and 57 had RFA alone. Recurrences were lowest with resection: 52% versus 64% for RFA and resection, and 84% for the RFA alone. Liver-only recurrence after RFA was 44%. Four-year survival was 65% for resection, 36% for resection and RFA, and 22% for RFA alone.[91] A multivariate analysis of these patients found the type of procedure (i.e., resection, RFA and resection, or RFA alone) influenced survival, with the RFA-alone group having the lowest survival. Of course, RFA was usually a component of therapy when resection was not possible, especially in cases where the anatomic distribution of tumors made complete resection impossible. Therefore, this is not a true comparison of RFA versus resection. However, 3-year survival for patients with one metastasis treated by resection or RFA was 80% versus 40%, respectively, suggesting RFA cannot replace resection.

The combination of RFA and HAI may be useful. In an M.D. Anderson study on 50 patients, 32% remained tumor-free at a 20-month median followup. Recurrence at the site of RFA was seen in 10% and new liver metastases in 30%.[246] Kainuma and coworkers treated nine patients with bilobar disease with RFA and regional chemotherapy with 5-FU, doxorubicin, and cisplatin.[247] The local recurrence rate was 55%, and 2-year survival was 39%. Martin and colleagues treated 21 patients with RFA and HAI FUDR. With a median followup time of 24 months, the median survival is 30 months.[248]

RFA has been used in other tumor types. In a Chinese series on 240 patients with HCC and 44 patients with liver metastases treated with RFA, patients with liver metastases had a higher extrahepatic recurrence (P = 0.019) and shorter DFS (P = 0.007). Patients with multiple liver metastases had a higher local and extrahepatic recurrence[249] in neuroendocrine tumors. RFA can relieve symptoms in 95% of patients. [251] [252] Livraghi and associates reported on 24 breast cancer patients whose liver lesions were treated with RFA, and 10 patients are still alive and free of disease with a median followup time of 10 months.[252]

Toxicity from RFA is clearly outlined in a series on 312 patients that included liver abscesses (7), portal vein thrombosis (3), pleural effusion (5), colon perforation (1), and renal insufficiency (1).[253]

In conclusion, RFA may have a role during surgical resection when one side of the liver is resected and small disease exists on the other side that cannot be resected. In patients who have undergone a resection and develop a small recurrence, percutaneous or laparoscopic RFA can be used if the lesion can be reached easily and is not close to large vessels. The presence of blood vessels near the tumors causes conduction of thermal energy away from the tumor and spares killing the tumor near the blood vessel.[254] Whether these techniques are more useful than chemotherapy, or should be used in combination with chemotherapy, are not known; the EORTC are exploring the use of RFA and chemotherapy versus chemotherapy alone.

PERCUTANEOUS ETHANOL INJECTION

PEIT was first performed in 1983 in Japan. Ultrasound guidance is used to place up to 30 mL of absolute ethanol into the lesion.[255] In patients with primary hepatoma, this treatment produced 5-year survivals of 43% for small lesions. Suzuki and coworkers[256] assigned 42 patients with HCC less than 3 cm to three groups—(1) chemolipiodolization, (2) chemolipiodolization followed by gelatin sponge transcatheter embolization (TAE) or (3) PEIT—and demonstrated a decrease in local recurrence with PEIT. Local recurrences at 1 year were 61%, 29%, and 20% for groups 1, 2, and 3, respectively. Shiina and colleagues[257] reported a 10-year survival of 66% using PEIT on single hepatocellular lesions smaller than 2 cm. In pooled data on 11,000 patients with HCC from Japan, 3-year survival for surgical resection, PEIT, or embolization was 58%, 53%, and 20%, respectively. Other researchers, however, report a higher recurrence rate after PEIT as compared with surgical resection. The size of the lesion also affects outcome. In an Italian study on 26 patients with metastatic disease, 13 of 15 patients with lesions smaller than 2 cm had responses, whereas among the 6 patients with lesions larger than 4 cm, no response was seen.[257] Yamamoto and associates[258] randomized 100 patients to TAE versus TAE + PEIT. The 3-year survival was 20% for the TAE group and 50% for the TAE + PEIT group, respectively (P = 0.05). This technique was also useful for treating small neuroendocrine tumors, possibly because they are highly hypervascular. At present, the technique needs further study to determine where it fits into the therapeutic armamentarium and whether it will increase survival for patients with metastatic liver tumors. Pending further study, it seems that PEIT and cryosurgery could be applicable to patients with small metastatic lesions who cannot undergo surgical resection. Whether these techniques will be more beneficial than regional hepatic arterial therapy is not clear, because they only treat visible disease and do not deal with possible small metastases that are not visible.

ISOLATION PERFUSION

To administer high drug concentrations locally, the liver can be isolated by clamping the hepatic arteries, vena cava, and portal vein and then placing a catheter in the hepatic artery to perfuse the liver. A catheter in the retrohepatic vena cava drains the liver, and extracorporeal filters allow removal of chemotherapeutic agents and simplify the technique of isolated perfusion. With a double-balloon inferior vena cava catheter, doses as high as 5,000 mg/m2 of 5-FU and 120 mg/m2 of doxorubicin have been administered. An initial trial using 30 mg/m2 of mitomycin C produced venoocclusive disease in four of the nine patients.[259] With the use of melphalan (L-PAM), toxicity was decreased, and with doses of 30 mg/kg, complete responses were seen in 2 of 9 patients.[260] Hyperthermic isolation perfusion of tumor necrosis factor and melphalan is being investigated. Melphalan (1.5 mg/kg) and tumor necrosis factor (1 mg) over 60 min of hyperthermic infusion produced a 75% response rate.[261]

GENE THERAPY

Tumors largely restricted to the liver (primary or metastatic) can potentially be treated by gene therapy. The gene transfer agents can be injected locally or via the hepatic artery. One trial involves the use of an adenovirus vector carrying the wild-type p53 gene. Alteration in p53 function is present in more than half of all malignancies, and reexpression of wild-type p53 can result in apoptosis and in tumor shrinkage in rodents. A phase I study of recombinant adenovirus encoding wild-type p53 administered via the hepatic artery produced no responses in 19 patients.[262] Transgene expression in tumor tissue was seen in patients receiving the highest dose levels. Other gene therapies involve the use of prodrug genes to convert innocuous drugs into active chemotherapeutic agents (cytosine deaminase converts 5-fluorocytosine to 5-FU).[263]

To deliver directed immunotherapy, Rubin and coworkers[264] injected HLA-B7 gene on a liposomal vector into liver tumors. No responses were seen in 15 patients in the phase I study, but plasmid DNA was detected in 14 of 15 patients. Another concept being evaluated is to have cells express a gene such as thymidine kinase and then kill the cells by use of a ganciclovir, which is converted by thymidine kinase to an active metabolite.[263]

RADIATION THERAPY

In this section, the evolving role of radiation therapy for liver metastases will be reviewed. The first experience in using radiation therapy for liver metastases was with external-beam whole-liver radiotherapy (with and without systemic or regional chemotherapy), in which the doses that could be delivered safely were not high enough to eradicate metastases. More recently, technologic advances in radiation treatment planning allow high-dose radiation therapy to be delivered conformally around liver metastases safely, with the potential for eradication of disease and cure, in appropriately selected patients. Other types of radiation have also been used to treat liver metastases including brachytherapy and hepatic arterial delivery of 90Y-tagged microspheres. Here, the experience following whole-liver irradiation, conformal radiation therapy, brachytherapy, and hepatic arterial 90Y for treatment of liver metastases will be reviewed.

Whole-Liver Irradiation

The approach of using whole-liver irradiation for metastases is limited by the low tolerance of the whole liver to irradiation, with doses required to be less than 30 Gy over 3 weeks to avoid liver toxicity. Several clinical trials of whole-liver irradiation for liver metastases have established the safe whole-liver doses that can be delivered in a variety of fractionations. [266] [267] The duration of response and survival rates tend to be short [268] [269] [270] ( Table 59-26 ). In contrast, low-dose whole-liver irradiation can produce palliation of painful liver metastases in the majority of cases. [267] [270] [271]


Table 59-26   -- Results of Treatment of Metastatic Cancer to the Liver Treated with Whole-Liver Irradiation Alone

Reference

Histology[*]

Dose (Gy/No. of Fractions)

No. of Patients

Response (% Total)

Median Survival (mo)

Hepatitis[†] Toxicity

Borgelt et al[266]

38% Colorectal

21–30/7–19

103

55[‡]

3

0

RTOG 76-05

 

 

 

 

 

 

Leibel et al[269]

48% Colorectal

21/7 (± misonidazole)

187

80[‡]/7[§]

4

0

RTOG 80-03

 

 

 

 

 

 

Phillips et al[267]

56% GI

≈20–37.5/8

36

72[‡]

ND

1

Prasad et al[268]

33% Colorectal

≈25/16

27

70[‡]

4

0

Russell et al[265]

60% Colorectal

27/15

53

 

4

0

RTOG 84-05

 

30/20

69

ND

4

0

 

 

33/22

51

 

4

2

GI, gastroimestinal; ND, not determined.

 

*

Predominant histology.

Number of patients with ≥ grade 3 radiation hepatitis.

Subjective decrease in pain.

§

Objective (CT scan).

 

In an attempt to improve on the poor outcomes following whole-liver irradiation alone, whole-liver irradiation has been combined with systemic or regional chemotherapy. The most widely used drugs in this effort have been the fluoropyrimidines because of their activity against CRC and their radiation-sensitizing properties. [272] [273] [274] The results of some of these trials are summarized in Table 59-27 . In general, the response rates and survival rates following combined-modality therapy seem to be superior to those obtained following whole-liver irradiation alone. Selection bias may be at least partially responsible for this effect. A recent study demonstrated that whole-liver radiation (20 Gy in 10 fractions) does not improve the efficacy of 5-FU for patients who have diffuse liver metastases.[274]


Table 59-27   -- Results of Treatment of Metastatic Cancer to the Liver Treated with Whole-Liver Irradiation with Chemotherapy

Reference

Dose (Gy/Fractions)

Chemo

Route

No. of Patients

Response (% Total)

Median Survival (mo)

Hepatic[*] Toxicity

Ajlouni et al[309]

21–30/14–20

FUDR

IAH

10

30[†]

9

0

Byfield et al[310]

15–30/12[‡]

FUDR

IAH

28

ND

9

1

Friedman et al[311]

13.5–21/5–7

5-FU, Dox

IAH

22

48[†]

3

1

Herbsman et al[312]

25–30/15

FUDR

IAH

13

70[§]

16

0

Lawrence et al[313]

33/22

FUDR

IAH

19

39[†]

7

0

 

36/24

 

 

13

ND

ND

3

Lokich et al[314]

19.5–30/10–12

5-FU or FUDR

IAH

12

63[§]

ND

0

McCracken et al[206]

19.5/13

5-FU, Mito C

IAH

13

(adjuvant)

ND

1

Raju et al[315]

21/1.5

FUDR or FU

IAH or IV[¶]

12

83[§]

14

0

Rotman et al[316]

≈22.5–32.3/15

5-FU

IV

27

83[§]

6

0

Sherman et al[317]

15–30/7–10

5-FU or Pro ± HU Cy ± 5-FU

IV

50[#]

90[§]

4

0

Webber et al[318]

25/10

FUDR

IAH

25

72[§]

12

0

Wiley et al[319]

25.5/17

5-FU

IAH

19

37[†]

6

0

Volberding et al[320]

21/7

5-FU, Dox, MTX

IAH

27

33[†]

7

0

Cy, cyclophosphamide; Dox, doxorubicin; 5-FU, 5-fluorouracil; FUDR-fluorodeoxyuridine; HU, hydroxyurea; IAH, intra-arteria hepatic infusion; IV, intravenous infusion; Mito C, mitomycin C; MTX, methotrexate; ND, not determined; Pro, procarbazine.

 

*

Number of patients with ≥ grade 3 radiation hepatitis.

Objective response (CT or radionuclide scan documenting 50% decrease in bidimensional proudct).

Split-course therapy.

§

Subjective response (e.g., decrease in pain).

FUDR (IAH) in four patients, 5-FU IV in eight patients.

#

Includes 19 patients who received RT only.

 

Conformal Radiation Therapy

Technical advances in liver cancer imaging, radiation planning, methods to account for breathing motion during radiation delivery and image guidance at the time of radiation delivery (Figs. 59-13 and 59-14 [13] [14]) have made it possible to deliver high-dose radiation safely to focal liver metastases, while sparing irradiation of the uninvolved liver, using a variety of radiation fractionation schedules.[275] In most studies, higher doses of radiation have been associated with more durable local control rates than lower doses, regardless of the fractionation schedule. [277] [278] [279]

 
 

Figure 59-13  Planning CT showing multiple liver metastases (in blue), treated with multiple radiation beams. Insert is a cutout of a verification CT obtained in the radiation treatment position immediately before therapy, registered to the planning CT scan.

 

 

 
 

Figure 59-14  Conformal radiation dose distribution, with the treatment isodose 57 Gy in six fractions, conforming to the target volume (pink).

 

 

Since the late 1980s at the University of Michigan, a series of phase I/II trials for patients with unresectable intrahepatic cancer have investigated dose-escalated conformal radiation therapy delivered concurrently with hepatic arterial chemotherapy (predominantly floxuridine [0.2 mg/kg/day]). In one of the first studies, the objective response rate of 22 patients with unresectable CRC liver metastases, treated with as much as 72.6 Gy at 1.5 Gy twice daily, was 50% (2 complete remission, 9 partial response, 11 stable disease),[279] with a median survival of 20 months. Similar results were obtained in subsequent studies,[280] including the most recent study in which the prescription dose was individualized based on the volume of liver irradiated and risk of toxicity, allowing higher doses (as much as 90 Gy at 1.5 Gy twice daily) to be delivered safely to more patients. The median survival of 47 patients with liver metastases (median diameter ∼10 cm) treated on this study was 17.2 months.[277]

Stereotactic body radiation therapy (SBRT), referring to a limited number of high-dose fractions delivered very conformally to targets, using biologic doses of radiation higher than those used in standard fractionation, has also been used to treat liver metastases. [279] [282] [283] [284] [285] [286] [287] [288] Safety of 1- to 10-fraction SBRT has been described in several retrospective series and more recently confirmed in prospective dose escalation studies ( Table 59-28 ). Blomgren and colleagues from Sweden first reported a response rate of 43% for 14 liver metastases treated with 20 to 45 Gy in one to four fractions,[262] with a prolonged time to maximal response (e.g., maximal response at 16 months for a 13-cm liver metastases). No liver toxicity was seen in patients with metastases, but hemorrhagic gastritis was seen in one patient. In an update in 1998, the local control rate was 95% with a mean survival of 17.8 months for 21 liver metastases.[288] SBRT (20 Gy × 2 or 15 Gy × 3) has also been used safely in patients with recurrent liver metastases following hepatic resection for CRC metastases, with no serious toxicity and local control 13 to 101 months following surgery.[289]


Table 59-28   -- Selected Results of Liver Metastases Treated with Conformal or Stereotactic Radiation Therapy

 

 

TUMOR TYPE

 

LOCAL CONTROL

SURVIVAL

Reference

No. of Patients

Mets

Other

Dose, No. of Fractions

%

Time

%

Time

Ben-Josef et al[277]

47

47+

 

<90 Gy, <60

95

 

 

17. 2 mo median

Blomgren et al[281]

23

14

9

20–45 Gy, 1–4

NR

 

 

NR

Herfarth et al[290]

37

60

 

15–26 Gy, 1

81

18 mo

 

 

Wulf et al[291]

23

23

 

30 Gy, 3

61

24 mo

41

24 mo

Schefter et al[278]

18

25

1

36–60 Gy, 3

NR

 

 

NR

Mendez Romero et al[286]

25

34

11

37.5 Gy, 3

82

24 mo

50

24 mo

Katz et al[285]

69

174

 

30–55 Gy, 5–15

57

20 mo

37

20 mo

NR, not reported.

 

 

 

A prospective study of escalated single-fraction SBRT (14 Gy to 26 Gy) did not find a maximal tolerated dose in 60 liver tumors (56 metastases) with a median tumor size of 10 mL (1–132 mL) and found an actuarial local control rate of 81% at 18 months following SBRT. [282] [291] SBRT delivered in three fractions (37.5 Gy total) has also been reported to be safe in small liver metastases, with 2-year local control and survival rates of 61% and 41%, respectively. [283] [292] A North American prospective study confirmed the safety of three-fraction SBRT in 18 patients with 25 tumors of maximal diameter 6 cm.[278] A Canadian prospective study has shown the feasibility of delivering six-fraction SBRT using an individualized dose allocation approach as first described by the Michigan group, for liver cancers ranging from 3 to 3,000 mL.[292]

More recently, outcomes following SBRT for 174 liver metastases from colorectal, pancreatic, breast, and lung cancer in 69 patients were reported.[285] The median dose delivered was 48 Gy (range 30–55 Gy) at 2 to 6 Gy per fraction. The local control was 76% and 57% at 10 and 20 months, respectively, with an overall medial survival of 14.5 months. No grade 3 toxicity was reported. Based on this experience, 10-fraction SBRT is being studied in a Radiation Therapy Oncology Group study that is now open.

Brachytherapy

High-dose radiation can also be delivered to focal liver metastases, with maximal sparing of dose to uninvolved liver, using interstitial brachytherapy, in which radiation is delivered from radioactive sources placed within or near the tumor. [294] [295] [296] [297] [298] High-dose-rate iridium-192 afterloaded to applicators placed at the time of laparotomy was used to deliver 20 to 30 Gy in 2 to 13 settings, with an actuarial local control rate of 26% at 26 months and two complete responses seen in lesions subsequently biopsied.[296] Iodine-125 seed implants, which deliver low-dose-rate irradiation (less than ∼0.15 Gy/hr) over several months also have been used to treat liver metastases, with 10 of 11 tumors controlled at 1 year.[295] More recently, CT-guided placement of iridium-192 for liver metastases has been shown to be feasible in a phase II study[294] of 20 patients with liver cancers (19 metastases, 1 cholangiocarcinoma) unsuitable for thermal ablation. The mean tumor diameter was 7.7 cm (5.5–10.8 cm) for peripheral lesions and 3.6 cm (2.2–4.9 cm) for hilar cancers. The dose delivered ranged from 12 to 25 Gy. Two serious complications were observed. One patient had an intrahepatic hemorrhage on removal of the brachytherapy sources. Another patient developed obstructive jaundice 14 days after brachytherapy and subsequent elevated bilirubin and liver failure 9 months later, perhaps associated with biliary injury from high-dose radiation. Mild increases in liver enzymes and bilirubin without clinical symptoms were common. The 1-year tumor control rates were 40% and 71% for the large peripheral and small hilar lesions, respectively, with an overall 1-year survival rate of 83%.[294] Interstitial brachytherapy is a highly specialized method of delivering very conformal radiation for liver metastases, only available at a few centers worldwide.

Hepatic Arterial Radioisotopes

Another method of delivering high-dose radiation to liver metastases is by infusing radiolabeled glass or resin microspheres into the hepatic artery (sometimes referred to as selective internal radiation therapy). Although a variety of radioisotopes have been investigated in this setting, the available commercial products for liver cancer use 90Y. 90Y is a pure β-emitter with a half-life of 64.5 hr and an average electron range of approximately 2.5 cm and an effective pathlength of 5.3 mm (meaning that 90% of the energy is deposited within a 5.3-mm radius of the microsphere). 90Y is incorporated into stable glass microspheres (TheraSphere, MDS Nordion Inc., Ottawa, Canada) or resin-based microspheres (SIRTex Medical, Inc., Lane Cove, New South Wales, Australia) which are infused into the hepatic artery as a form of regional therapy for well-vascularized liver tumors. [299] [300] [301] [302] [303] [304] [305] A typical prescribed dose is 120 to 150 Gy, where the microspheres are primarily deposited at the periphery of the metastases.[305] A reduction in CEA and radiographic responses are common following hepatic arterial 90Y microspheres for liver metastases. Hepatic arterial 90Y microspheres have been combined safely with regional 5-FU.[300]

A randomized trial comparing the combination of hepatic arterial 90Y microspheres and hepatic arterial floxuridine to hepatic arterial floxuridine alone was conducted in 74 patients with unresectable liver confined colorectal metastases who underwent implantation of an HAI pump and chronic floxuridine in 12-day cycles at 4-week intervals. The patients were randomized to receive a single dose of 90Y microspheres through the hepatic artery within 4 weeks of surgery versus no other therapy. The overall response rate and median time to progression were increased in the group receiving 90Y microspheres (44% vs. 18% [P < 0.01] and 16 months vs. 10 months [P < 0.001], respectively), with a trend toward an improvement in overall survival (17% vs. 7% 3-year survival).[299] This randomized trial demonstrated that irradiation of metastases improves outcomes over hepatic arterial chemotherapy alone. However, the response rate for hepatic arterial floxuridine in this study was substantially less than typically reported (see earlier in this chapter), and confirmatory studies are warranted.

Challenges associated with hepatic arterial delivery of 90Y include the lack of understanding of dosimetry, the requirement for specialized interventional radiology expertise, and the potential for a radiation hazard associated with lost radioisotopes.

Liver Tolerance to Irradiation

In the early 1960s it was found that doses greater than 30 Gy at 2 Gy per fraction to the whole liver led to an unacceptable risk of liver toxicity. Eight fractions of 2.25 Gy to the whole liver was found to be safe, but a small increase in fraction size to 3.5 Gy was reported in 1973 to be associated with an unacceptable rate of liver toxicity (8 of 25 patients). [307] [308] The tolerance of the liver to whole-organ irradiation does not seem to be substantially altered by the concomitant use of fluoropyrimidines. In contrast, whole-liver irradiation in combination with alkylating agents or mitomycin C is associated with an increased risk of liver toxicity. [309] [310] [311] Liver toxicity following irradiation has historically been referred to as “radiation hepatitis” and more recently termed “radiation-induced liver disease” (RILD),[308] because there is no evidence of hepatitis on pathologic examination. This complication is a clinical syndrome consisting of anicteric ascites and painful hepatomegaly, occurring in the absence of disease progression, usually within 3 months following a course of radiation therapy. Laboratory evaluation demonstrates a marked elevation of alkaline phosphatase out of proportion to the modest increases in the transaminases. Although the majority of patients recover from RILD, it may progress to liver failure and death.[308] The pathophysiology of RILD is not well understood. Pathologically, venoocclusive disease, similar to that seen following bone marrow translation, is seen.

Partial liver radiation therapy was first reported in 1965 by Ingold and associates, who safely delivered as much as 55 Gy to parts of the liver.[311] Others confirmed that high-dose radiation therapy could be delivered safely, as long as a substantial portion of the normal liver was spared. Conformal radiation planning permits portions of the liver to be treated with doses of radiation far higher than what the entire liver can tolerate so long as a sufficient volume of uninvolved liver can be spared from irradiation, similar to how a surgeon can resect a substantial fraction of the liver if the remaining liver is functional. Conformal radiation treatment planning allows the fraction of uninvolved liver irradiated to be quantified. Theoretical models have been proposed to estimate the volume dependence of normal tissue tolerance—referred to as normal tissue complication probability (NTCP) models. [313] [314] [315] [316] Such an NTCP model has been used to describe the partial liver tolerance of 203 patients treated with conformal hyperfractionated radiation therapy and hepatic arterial floxuridine (17 of whom developed RILD). This analysis found that that mean liver dose can provide an estimate of the risk of RILD occurring, with a 5% risk of toxicity following 32 Gy and 37 Gy in 1.5 Gy twice daily for patients with primary liver cancer and metastases, respectively.[316]

The partial volume tolerance of the liver to hypofractionation or SBRT has not been well established, partially in that the majority of clinical SBRT experience has not required large liver volumes to be irradiated. Guidelines used in SBRT planning include sparing of 30% and 50% of the liver from 12 Gy and 7 Gy, respectively, for three- and one-fraction SBRT),[283] and ensuring that at least 700 mL of uninvolved liver receives less than 15 Gy in three fractions.[278] Dawson and coworkers have shown the feasibility of using an NTCP model for allocation of six-fraction SBRT in over 80 patients,[292] with a wide range of liver volumes irradiated.

The safe delivery of high-dose 90Y to small volumes seems consistent with the partial volume estimates from conformal radiation and SBRT series, where the upper limit on the dose of radiation that can be delivered to an effective liver volume irradiated of 20% or less has not been established. The lack of a validated dose distribution in 90Y treatment makes partial liver tolerance analysis challenging for 90Y therapy.

Extrapolation of NTCP models and partial liver tolerances to different centers must be done with caution, because the results may not be valid for different patient populations treated with different treatments. The partial liver tolerance to irradiation, especially in diseased livers and following SBRT and 90Y, should be measured and validated in prospective studies.

FUTURE DIRECTIONS

Although it is becoming established that radiation therapy, delivered using conformal radiotherapy, SBRT, brachytherapy, or 90Y microspheres, can be used safely to treat liver metastases with the potential for sustained local control, recurrences outside the irradiated volume are not infrequent, providing a rationale for combining radiation therapy with other therapies. One possibility would be to combine radiation therapy with repeated cycles of modern hepatic arterial and systemic chemotherapy. More potent radiation tumor sensitizers could also be explored. Studies of targeted therapies combined with radiotherapy should be considered, given the radiation sensitization properties of many targeted agents and survival gains observed following the combination of radiation and targeted agents in other clinical sites.[317] Conversely, normal-tissue radiation protectors, as well as more technologic advances, may allow higher doses to be delivered to more tumors safely. For instance, the free-radical scavenger amifostine protects the normal liver (but not tumor) from radiation in preclinical studies.[318] It is hoped that these approaches will permit a greater fraction of patients to benefit from high-dose therapy and will increase local control in patients with localized unresectable intrahepatic cancers.

CONCLUSIONS

Liver metastases, especially from colorectal primaries, are treatable and potentially curable. Imaging techniques such as CT, MRI, and sonography have advanced in recent years and have led to increased sensitivity and specificity in the diagnosis of liver metastases. When properly applied, the techniques are nearly as sensitive as surgical exploration in the detection of hepatic lesions and in differentiating benign from malignant processes.

Liver surgery has improved greatly in the past two decades. Dissections along nonanatomic lines have permitted the resection of multiple lesions that might previously have been considered unresectable. The new technique of vascular exclusion might improve the safety of major hepatic resections. Resection of solitary hepatic metastases or up to two to four metastases from CRC should be regarded as the best treatment for this condition.

Noninvasive approaches to liver metastases such as cryosurgery, chemoembolization, radiofrequency ablation, or alcohol injections are available. The usefulness of these techniques compared with surgery or regional therapy is being investigated.

For focal liver metastases unsuitable for or refractory to surgery or other ablative approaches, external-beam conformal radiation therapy, SBRT, or brachytherapy may be used to attempt to locally control disease. Hepatic arterial 90Y-irradiation can be used safely to treat patients with diffuse metastases, with normal bilirubin levels. The optimal timing of radiation and chemotherapy has not been established, and there is rationale for investigating the combination of simultaneous regional or systemic therapies with radiation therapy in future studies.

REFERENCES

  1. Kalender W, Polacin A: Physical performance characteristics of spiral CT scanning.  Med Phys1991; 18:910-915.
  2. Baron RL, Moss AA: The liver.   In: Moss AA, Genant HK, ed. Computed Tomography of the Body with Magnetic Resonance Imaging,  Philadelphia: WB Saunders; 1992:735-821.
  3. Hoffman M, Shi H, Schmitz B, et al: Noninvasive coronary angiography with mulitslice computed tomography.  JAMA2005; 293:2471-2478.
  4. Phillips VBM: The liver and spleen.   In: Putnam DRC, ed. Diagnostic Imaging,  Philadelphia: WB Saunders; 1988:937.
  5. Bipat S, Leeuwen MV, Comans E, et al: Colorectal liver metastases: CT, MR imaging, and PET for diagnosis—meta-analysis.  Radiology2005; 237:123-131.
  6. Albrecht T, Blomley M, Burns P, et al: Improved detection of hepatic metastases with pulse-inversion US during the liver-specific phase of SHU 508A: multicenter study.  Radiology2003; 227:361-370.
  7. Krix M, Kiesslink F: Low mechanical index contrast-enhanced ultrasound better reflects high arterial perfusion of liver metastases than arterial phase computed tomography.  Invest Radiol2004; 39:216-222.
  8. Brannigan M, Burns P, Wilson S: Blood flow patterns in focal liver lesions at microbubble-enhanced US.  Radiographics2004; 24:921-925.
  9. Machi J, Isomoto H, Yamashita Y, et al: Intraoperative ultrasonography in screening for liver metastases from colorectal cancer: comparative accuracy with traditional procedures.  Surgery1987; 101:678-684.
  10. Russo A, Sparacino G, Plaja S, et al: Role of intraoperative ultrasound in the screening of liver metastases from colorectal carcinoma: initial experiences.  J Surg Oncol1989; 42:249-255.
  11. Wiering B, Krabbe P, Gerrit G, et al: The impact of fluoro-18-deoxyglucose-positron emission tomography in the management of colorectal liver metastases.  Cancer2005; 104:2658-2670.
  12. Ruers TJ, Langenhoff BS, Neeleman N, et al: Value of positron emission tomography with [F-18]fluorodeoxyglucose in patients with colorectal liver metastases: a prospective study.  J Clin Oncol2002; 20:388-395.
  13. Lai DT, Fulham M, Stephen MS, et al: The role of whole-body positron emission tomography with [18F]fluorodeoxyglucose in identifying operable colorectal cancer metastases to the liver.  Arch Surg1996; 131:703-707.
  14. Rydzewski B, Dehdashti F, Gordon BA, et al: Usefulness of intraoperative sonography for revealing hepatic metastases from colorectal cancer in patients selected for surgery after undergoing FDG PET.  Am J Roentgenol2002; 178:353-358.
  15. Vitola JV, Delbeke D, Meranze SG, et al: Positron emission tomography with F-18-fluorodeoxyglucose to evaluate the results of hepatic chemoembolization.  Cancer1996; 78:2216-2222.
  16. Tanaka T, Kawai Y, Kanai M, et al: Usefulness of FDG-positron emission tomography in diagnosing peritoneal recurrence of colorectal cancer.  Am J Surg2002; 184:433-436.
  17. Strasberg S, Dehdashti F, Siegel B, et al: Survival of patients evaluated by FDG-PET before hepatic resection for metastatic colorectal carcinoma: a prospective database study.  Ann Surg2001;233-293.
  18. Kalff V, Hicks RJ, Ware RE, et al: Evaluation of high-risk melanoma: comparison of [18F]FDG PET and high-dose 67Ga SPET.  Eur J Nucl Med Mol Imaging2002; 29:506-515.
  19. Fong Y, Fortner J, Sun RL, et al: Clinical score for predicting recurrence after hepatic resection for metastatic colorectal cancer: analysis of 1001 consecutive cases.  Ann Surg1999; 230:309-318.discussion 18–21
  20. Fernandez F, Drebin J, Jeffrey D, et al: Five-year survival after resection of hepatic metastases from colorectal cancer in patients screened by positron emission tomography with F-18 fluorodeoxyglucose (FDG-PET).  Ann Surg2004; 240:447-450.
  21. Wernecke K, Rummeny E, Bongartz G, et al: Detection of hepatic masses in patients with carcinoma: comparative sensitivities of sonography, CT, and MR imaging.  Am J Roentgenol1991; 157:731-739.
  22. Kemeny MM, Sugarbaker PH, Smith TJ, et al: A prospective analysis of laboratory tests and imaging studies to detect hepatic lesions.  Ann Surg1982; 195:163-167.
  23. Vassiliades VG, Foley WD, Alarcon J, et al: Hepatic metastases: CT versus MR imaging at 1.5T.  Gastrointest Radiol1991; 16:159-163.
  24. Hussain S, Semelka R: Hepatic imaging: comparison of modalities.  Radiol Clin North Am2005; 43:929-947.
  25. Braga L, Gulle U, Semelka R: Pre, peri-, and posttreatment imaging of liver lesions.  Radiol Clin North Am2005; 43:915-927.
  26. Voroney JJ, Brock KK, Eccles C, et al: Prospective comparison of CT and MRI for liver cancer delineation using deformable image registration.  Int J Radiat Oncol Biol Phys2006; 66:780-791.
  27. Dodd 3rd GD, Miller WJ, Baron RL, et al: Detection of malignant tumors in end-stage cirrhotic livers: efficacy of sonography as a screening technique.  Am J Roentgenol1992; 159:727-733.
  28. Delbeke D, Vitola JV, Sandler MP, et al: Staging recurrent metastatic colorectal carcinoma with PET.  J Nucl Med1997; 38:1196-1201.
  29. Ying M, Martin D, Karabulut N, et al: Comparison of MR and PET imaging for the evaluation of liver metastases.  J Magn Reson Imaging2003; 17:343-349.
  30. Takahashi S, Kuroki Y, Nasu K, et al: Positron emission tomography with F-18 fluorodeoxyglucose in evaluating hepatic metastases down staged by chemotherapy.  Anticancer Res2006; 26:4705-4711.
  31. Ziessman HA, Silverman PM, Patterson J, et al: Improved detection of small cavernous hemangiomas of the liver with high-resolution three-headed SPECT.  J Nucl Med1991; 32:2086-2091.
  32. Vassiliades VG, Bernardino ME: Percutaneous renal and adrenal biopsies.  Cardiovasc Intervent Radiol1991; 14:50-54.
  33. Tartter PI, Slater G, Gelernt I, Aufses A: Screening for liver metastases from colorectal cancer with carcinoembryonic antigen and alkaline phosphatase.  Ann Surg1981; 193:357-360.
  34. Ranson J, Adams P, Localio S: Preoperative assessment for hepatic metastases in carcinoma of the colon and rectum.  Surg Gynecol Obstet1973; 137:435-438.
  35. Kemeny M, Hogan J, Ganteaume B, et al: Preoperative staging with computerized axial tomography and biochemical laboratory tests in patients with hepatic metastases.  Ann Surg1986; 203:169-172.
  36. Doroshow J, Multhauf P, Leong L, et al: Prospective randomized comparison of fluorouracil versus fluorouracil and high-dose continuous infusion leucovorin calcium for the treatment of advanced measurable colorectal cancer in patients previously unexposed to chemotherapy.  J Clin Oncol1990; 8:491-501.
  37. Iwatsuki S, Sheehan DG, Starzl TE: The changing face of hepatic resection.  Curr Probl Surg1989; 26:281-379.
  38. Ayav A, Jiao LR, Habib NA: Bloodless liver resection using radiofrequency energy.  Dig Surg2007; 24:314-317.
  39. Stephen MS, Sheil AG, Thompson JF, et al: Aortic occlusion and vascular isolation allowing avascular hepatic resection.  Arch Surg1990; 125:1482-1485.
  40. Delva E, Camus Y, Nordlinger B, et al: Vascular occlusions for liver resections—operative management and tolerance to hepatic ischemia: 142 cases.  Ann Surg1989; 209:211-218.
  41. Bismuth H, Castaing D, Garden J: Major hepatic resection under total vascular exclusion.  Ann Surg1989; 210:13-19.
  42. Zulim RA, Rocco M, Goodnight Jr JE, et al: Intraoperative autotransfusion in hepatic resection for malignancy. Is it safe?.  Arch Surg1993; 128:206-211.
  43. Fujimoto J, Okamoto E, Yamanaka N, et al: Efficacy of autotransfusion in hepatectomy for hepatocellular carcinoma.  Arch Surg1993; 128:1065-1069.
  44. Giakoustidis E, Drosinopoulos P, Agouridakis K, Galanis N: Surgical treatment of liver injuries by application of Fibrinkleber.  World J Surg1985; 9:144-148.
  45. Fong Y, Brennan M, Brown K, et al: Drainage is unnecessary after elective liver resection.  Am J Surg1996; 171:158-162.
  46. Burt BM, Brown K, Jarnagin W, et al: An audit of results of a no-drainage practice policy after hepatectomy.  Am J Surg2002; 184:441-445.
  47. Miyazaki M, Hiroshi I, Nakagawa K, et al: Hepatic resection of liver metastases from gastric cancer.  Am J Gastroenterol1997; 92:490.
  48. Harrison L, Brennan M, Newman E, et al: Hepatic resection for noncolorectal, nonneuroendocrine metastases: a 15-year experience with 96 patients.  Surgery1997; 121:625-632.
  49. Adam R, Aloia T, Figueras J, et al: Liver Met Survey: analysis of clinicopathologic factors associated with the efficacy of preoperative chemotherapy in 2,122 patients with colorectal liver metastases. In 2006 ASCO Annual Meeting. Atlanta, Georgia, USA, Am Soc Clin Oncol, 2006.
  50. Pocard M, Pouillart P, Asselain B, Salmon R: Hepatic resection in metastatic breast cancer: results and prognostic factors.  Eur J Surg Oncol2000; 26:155-159.
  51. Elias D, Lasser PH, Montrucolli D, et al: Hepatectomy for liver metastases from breast cancer.  Eur J Surg Oncol1995; 21:510-513.
  52. McEntee GP, Nagorney DM, Kvols LK, et al: Cytoreductive hepatic surgery for neuroendocrine tumors.  Surgery1990; 108:1091-1096.
  53. Que F, Nagorney D, Batts K, et al: Hepatic resection for metastatic neuroendocrine carcinomas.  Am J Surg1995; 169:36-43.
  54. Nordlinger B, Vaillant J, Guiguet M, et al: Survival benefit of repeat liver resections for recurrent colorectal metastases: 143 cases.  J Clin Oncol1994; 12:1491-1496.
  55. Elias D, Rougier P, Lasser P, et al: Reductional major surgery and chemotherapy in multimetastatic apudomas.  Ann Chir1988; 42:474-481.
  56. Cobourn CS, Makowka L, Langer B, et al: Examination of patient selection and outcome for hepatic resection for metastatic disease.  Surg Gynecol Obstet1987; 165:239-246.
  57. Wolf R, Goodnight J, Krag D, Schneider P: Results of resection and proposed guidelines for patient selection in instances of noncolorectal hepatic metastases.  Surg Gynecol Obstet1991; 173:454-460.
  58. Adam R, Chiche L, Aloia T, et al: Hepatic resection for noncolorectal nonendocrine liver metastases: analysis of 1,452 patients and development of a prognostic model.  Ann Surg2006; 244:524-535.
  59. Que F, Sarmiento JM, Nagorney DM: Hepatic surgery for metastatic gastrointestinal neuroendocrine tumors.  Cancer Control2002; 9:67-79.
  60. Pawlik TM, Vauthey JN, Abdalla EK, et al: Results of a single-center experience with resection and ablation for sarcoma metastatic to the liver.  Arch Surg2006; 141:537-544.
  61. DeMatteo RP, Shah A, Fong Y, et al: Results of hepatic resection for sarcoma metastatic to liver.  Ann Surg2001; 234:540-548.
  62. Sesto ME, Vogt DP, Hermann RE: Hepatic resection in 128 patients: a 24-year experience.  Surgery1987; 102:846-851.
  63. Lang H, Nussbaum KT, Kaudel P, et al: Hepatic metastases from leiomyosarcoma: a single-center experience with 34 liver resections during a 15-year period.  Ann Surg2000; 231:500-505.
  64. Rose DM, Essner R, Hughes TM, et al: Surgical resection for metastatic melanoma to the liver: the John Wayne Cancer Institute and Sydney Melanoma Unit experience.  Arch Surg2001; 136:950-955.
  65. Fong Y, Fortner J, Sun RL, et al: Clinical score for predicting recurrence after hepatic resection for metastatic colorectal cancer: analysis of 1001 consecutive cases.  Ann Surg1999; 230:309-321.
  66. Nordlinger B, Guigiet M, Vaillant JC, et al: Surgical resection of colorectal carcinoma metastases to the liver.  Cancer1996; 77:1254-1262.
  67. Taylor M, Forster J, Langer B, et al: A study of prognostic factors for hepatic resection for colorectal metastases.  Am J Surg1997; 173:467-471.
  68. Gayowski T, Iwatsuki S, Madariaga J, et al: Experience in hepatic resection for metastatic colorectal cancer: analysis of clinical and pathologic risk factors.  Surgery1994; 116:703-711.
  69. Jamison R, Donohur J, Nagorney D, et al: Hepatic resection for metastatic colorectal cancer results in cure for some patients.  Arch Surg1997; 132:505-511.
  70. Scheele J, Stang R, Altendorf-Hofmann A: Hepatic metastases from colorectal carcinoma: impact of surgical resection on the natural history.  Br J Surg1990; 77:1241-1246.
  71. Wilson S, Adson M: Surgical treatment of hepatic metastases from colorectal cancers.  Arch Surg1976; 111:330-334.
  72. Nordlinger B, Guiguet M, Vaillant JC, et al: Surgical resection of colorectal carcinoma metastases to the liver. A prognostic scoring system to improve case selection, based on 1568 patients. Association Francaise de Chirurgie.  Cancer1996; 17:1254-1262.
  73. Registry of Hepatic Metastases : Resection of the liver for colorectal carcinoma metastases: a multi-institutional study of indications for resection.  Surgery1988; 103:278-288.
  74. Hughes K, Scheele J, Sugarbaker P: Surgery for colorectal cancer metastatic to the liver.  Surg Clin North Am1989; 69:340-359.
  75. Hughes KS, Simon R, Songhorabodi S, et al: Resection of the liver for colorectal carcinoma metastases: a multi-institutional study of patterns of recurrence.  Surgery1986; 100:278-284.
  76. Fegiz G, Ramacciato G, Gennari L, et al: Hepatic resection for colorectal metastases: the Italian multicenter experience.  J Surg Oncol1991; 2:144-154.
  77. Doci R, Gennari L, Bignami P, et al: One hundred patients with hepatic metastases from colorectal cancer treated by resection: analysis of prognostic determinants.  Br J Surg1991; 78:797-801.
  78. Elias D, Lasser P, Rougier P, et al: Another failure in the attempt of definition of the indications to the resection of liver metastases of colorectal origin [French].  J Chir (Paris)1992; 129:59-65.
  79. Van Ooijen B, Wiggers T, Meijer S, et al: Hepatic resections for colorectal metastases in the Netherlands—a multi-institutional 10-year study.  Cancer1992; 70:28-34.
  80. Scheele J, Stangl R, Altendorf-Hofmann A, et al: Indicators of prognosis after hepatic resection for colorectal secondaries.  Surgery1991; 110:13-29.
  81. Fong Y, Blumgart L, Fortner J, et al: Pancreatic or liver resection for malignancy is safe and effective for the elderly.  Ann Surg1995; 222:426-434.
  82. Kornprat P, Jarnagin WR, Gonen M, et al: Outcome after hepatectomy for multiple (4 or more) colorectal metastases in the era of effective chemotherapy.  Ann Surg Oncol2007; 14:1151-1160.
  83. Adam R, Aloia T, Figueras J, et al: LiverMetSurvey: Analysis of clinicopathologic factors associated with the efficacy of preoperative chemotherapy in 2,122 patients with colorectal liver metastases. Proceedings of the 2006 ASCO Annual Meeting; Atlanta, Georgia, USA: American Society of Clinical Oncology; 2006.
  84. Adam R, Delvart V, Pascal G, et al: Rescue surgery for unresectable colorectal liver metastases downstaged by chemotherapy: a model to predict long-term survival.  Ann Surg2004; 240:644-658.
  85. Fernandez FG, Drebin JA, Linehan DC, et al: Five-year survival after resection of hepatic metastases from colorectal cancer in patients screened by positron emission tomography with F-18 fluorodeoxyglucose (FDG-PET).  Ann Surg2004; 240:438-450.
  86. Adson M, Van Heerden J, Adson M, et al: Resection of hepatic metastases from colorectal cancer.  Arch Surg1984; 119:647-651.
  87. Headrick JR, Miller DL, Nagorney DM, et al: Surgical treatment of hepatic and pulmonary metastases from colon cancer.  Ann Thorac Surg2001; 71:975-990.
  88. Fernandez-Trigo V, Shamsa F, Aldrete J, et al: Repeat liver resections from colorectal metastasis. Repeat Hepatic Resection Registry.  Cancer Treat Res1994; 69:185-196.
  89. Petrowsky H, Gonen M, Jarnagin W, et al: Second liver resections are safe and effective treatment for recurrent hepatic metastases from colorectal cancer: a bi-institutional analysis.  Ann Surg2002; 235:863-871.
  90. Fong Y, Blumgart L, Cohen A, et al: Repeat hepatic resections for metastatic colorectal cancer.  Ann Surg1994; 220:657-662.
  91. Abdalla EK, Vauthey JN, Ellis LM, et al: Recurrence and outcomes following hepatic resection, radiofrequency ablation, and combined resection/ablation for colorectal liver metastases.  Ann Surg2004; 239:818-827.
  92. Aloia T, Sebagh M, Plasse M, et al: Liver histology and surgical outcomes after preoperative chemotherapy with fluorouracil plus oxaliplatin in colorectal cancer liver metastases.  J Clin Oncol2006; 24:4983-4990.
  93. Bolton JS, Fuhrman GM: Survival after resection of multiple bilobar hepatic metastases from colorectal carcinoma.  Ann Surg2000; 231:743-751.
  94. Nordlinger B, Guiguet M, Vaillant JC, et al: Surgical resection of colorectal carcinoma metastases to the liver. A prognostic scoring system to improve case selection, based on 1568 patients. Association Française de Chirurgie.  Cancer1996; 77:1254-1262.
  95. Tanaka K, Shimada H, Matsuo K, et al: Outcome after simultaneous colorectal and hepatic resection for colorectal cancer with synchronous metastases.  Surgery2004; 136:650-659.
  96. Martin R, Paty P, Fong Y, et al: Simultaneous liver and colorectal resections are safe for synchronous colorectal liver metastasis.  J Am Coll Surg2003; 197:233-242.
  97. Weber JC, Bachellier P, Oussoultzoglou E, Jaeck D: Simultaneous resection of colorectal primary tumour and synchronous liver metastases.  Br J Surg2003; 90:956-962.
  98. Hao CY, Ji JF: Surgical treatment of liver metastases of colorectal cancer: strategies and controversies in 2006.  Eur J Surg Oncol2006; 32:473-483.
  99. Mentha G, Majno PE, Andres A, et al: Neoadjuvant chemotherapy and resection of advanced synchronous liver metastases before treatment of the colorectal primary.  Br J Surg2006; 93:872-878.
  100. Capussotti L, Muratore A: Neoadjuvant chemotherapy and resection of advanced synchronous liver metastases before treatment of the colorectal primary.  Br J Surg2006; 93:872-878.Br J Surg. 2006;93:1564; author reply.
  101. Kemeny M, Goldberg D, Beatty J, et al: Results of a prospective randomized trial of continuous regional chemotherapy and hepatic resection as treatment of hepatic metastases from colorectal cancer.  Cancer1986; 57:492-498.
  102. Wagman L, Kemeny M, Leong L, et al: A prospective randomized evaluation of the treatment of colorectal cancer metastatic to the liver.  J Clin Oncol1990; 8:1885-1893.
  103. Kemeny N, Huang Y, Cohen AM, et al: Hepatic arterial infusion of chemotherapy after resection of hepatic metastases from colorectal cancer.  N Engl J Med1999; 341:2039-2048.

103A. Kemeny NE, Gonen M: Hepatic arterial infusion after liver resection.  N Engl J Med  2005; 352:734-735.

  1. Kemeny MM: Updated data on survival. In Kemeny NE (ed): New York; 2005.
  2. Kemeny MM, Adak S, Gray B, et al: Combined-modality treatment for resectable metastatic colorectal carcinoma to the liver: surgical resection of hepatic metastases in combination with continuous infusion of chemotherapy—an intergroup study.  J Clin Oncol2002; 20:1499-1505.
  3. Lorenz M, Muller H-H, Schramm H, et al: Randomized trial of surgery versus surgery followed by adjuvant hepatic arterial infusion with 5-fluorouracil and folinic acid for liver metastases of colorectal cancer.  Surgery1998; 228:756-762.
  4. Lygidakis NJ, Sgourakis G, Vlachos L, et al: Metastatic liver disease of colorectal origin: the value of locoregional immunochemotherapy combined with systemic chemotherapy following liver resection. Results of a prospective randomized study.  Hepatogastroenterology2001; 48:1685-1691.
  5. Tono T, Hasuike Y, Ohzato H, et al: Limited but definite efficacy of prophylactic hepatic arterial infusion chemotherapy after curative resection of colorectal liver metastases: a randomized study.  Cancer2000; 88:1549-1556.
  6. Nonami T, Takeuchi Y, Yasui M, et al: Regional adjuvant chemotherapy after partial hepatectomy for metastatic colorectal carcinoma.  Semin Oncol1997; 24:56-130.
  7. Kusunoki M, Yanagi H, Noda M, et al: Results of pharmacokinetic modulating chemotherapy in combination with hepatic arterial 5-fluorouracil infusion and oral UFT after resection of hepatic colorectal metastases.  Cancer2000; 89:1228-1235.
  8. Asahara T, Kikkawa M, Okajima M, et al: Studies of postoperative transarterial infusion chemotherapy for liver metastasis of colorectal carcinoma after hepatectomy.  Hepatogastroenterology1998; 45:805-811.
  9. Kemeny N, Jarnigan W, Gonen M, et al: Phase I/II study of hepatic arterial therapy with floxuridine and dexamethasone in combination with intravenous irinotecan as adjuvant treatment after resection of hepatic metastases from colorectal cancer.  J Clin Oncol2003; 21:3303-3309.
  10. Kemeny NE, Jarnagin W, Gonen M, et al: Phase I trial of hepatic arterial infusion (HAI) with floxuridine (FUDR) and dexamethasone (DEX) in combination with systemic oxaliplatin (OXAL), fluorouracil (FU) + leucovorin (LV) after resection of hepatic metastases from colorectal cancer. In: 2005 ASCO Annual Meeting. Orlando, Florida, USA: American Society of Clinical Oncology; 2005.
  11. Alberts SR, Mahoney MR, Donohue JH, et al: Systemic capecitabine and oxaliplatin administered with hepatic arterial infusion (HAI) of floxuridine (FUDR) following complete resection of colorectal metastases (MCRC) confined to the liver: a North Central Cancer Treatment Group (NCCTG) phase II intergroup trial. In: 2006 ASCO Annual Meeting. Atlanta, Georgia, USA: American Society of Clinical Oncology; 2006.
  12. Langer B, Bleiberg H, Labianca R, et al: Fluorouracil (FU) plus l-leucovorin (l-LV) versus observation after potentially curative resection of liver or lung metastases from colorectal cancer (CRC): results of the ENG (EORTC/NCIC CTG/GIVIO) randomized trial. In: 2002 ASCO Annual Meeting Orlando, Florida, USA: American Society of Clinical Oncology; 2002.
  13. Ono T, Nagasue N, Kohno H, et al: Adjuvant chemotherapy with epirubicin and carmofur after radical resection of hepatocellular carcinoma: a prospective randomized study.  Semin Oncol1997; 26(Suppl 6):18-25.
  14. Lai E, Lo CM, Fan ST, et al: Postoperative adjuvant chemotherapy after curative resection of hepatocellular carcinoma.  Arch Surg1998; 133:183-188.
  15. Seifert JK, Springer A, Baier P, et al: Liver resection or cryotherapy for colorectal liver metastases: a prospective case control study.  Int J Colorectal Dis2005; 20:507-520.
  16. Pritchard KI: Liver metastases: can our understanding of their biology and prognostic value contribute to a strategy for optimum therapeutic management?.  Eur J Cancer1997; 33(Suppl 7):S11-S14.
  17. Nicholson G: Differential organ tissue adhesion, invasion and growth properties of metastatic rat mammary adenocarcinoma cells.  Breast Cancer Res Treat1988; 12:167.
  18. Nash CI, Jones SE, Moon T, et al: Prediction of outcome in metastatic breast cancer treated with adriamycin combination chemotherapy.  Cancer1980; 46:2380-2388.
  19. Carter S: Single and combination nonhormonal chemotherapy in breast cancer.  Cancer1972; 30:1543-1555.
  20. George SL, Hoogstraten B: Prognostic factors in the initial response to therapy by patients with advanced breast cancer.  J Natl Cancer Inst1978; 60:731-736.
  21. Dieras V, Chevallier B, Kerbrat P, et al: A multicentre phase II study of docetaxel 75 mg m-2 as first-line chemotherapy for patients with advanced breast cancer: report of the Clinical Screening Group of the EORTC. European Organization for Research and Treatment of Cancer.  Br J Cancer1996; 74:650-656.
  22. Kemeny N: The systemic chemotherapy of hepatic metastases.  Semin Oncol1983; 10:148-158.
  23. Kemeny N, Kemeny M, Lawrence TS: Liver metastases.   In: Abeloff AJ, Niederhuber JE, Kastan MB, McKenna WG, ed. Clinical Oncology,  3rd ed.. Philadelphia: Elsevier; 2004:1141-1178.
  24. Kemeny N, Yagoda A, Braun Jr D, Golbey R: Therapy for metastatic colorectal carcinoma with a combination of methyl-CCNU, 5-fluorouracil, vincristine and streptozotocin (MOF-Strep).  Cancer1980; 45:876-881.
  25. Tashiro T, Kawada Y, Sakurai Y, Kidani Y: Antitumor activity of a new platinum complex, oxalato (trans-l-1,2-diaminocyclohexane)platinum (II): new experimental data.  Biomed Pharmacother1989; 43:251-260.
  26. Mathe G, Kidani Y, Segiguchi M, et al: Oxalato-platinum or 1-OHP, a third-generation platinum complex: an experimental and clinical appraisal and preliminary comparison with cis-platinum and carboplatinum.  Biomed Pharmacother1989; 43:237-250.
  27. Saltz LB, Cox JV, Blanke C, et al: Irinotecan plus fluorouracil and leucovorin for metastatic colorectal cancer. Irinotecan Study Group.  N Engl J Med2000; 343:905-914.
  28. Douillard JY, Cunningham D, Roth AD, et al: Irinotecan combined with fluorouracil compared with fluorouracil alone as first-line treatment for metastatic colorectal cancer: a multicentre randomised trial.  Lancet2000; 355:1041-1047.
  29. Goldberg RM, Sargent DJ, Morton RF, et al: A randomized controlled trial of fluorouracil plus leucovorin, irinotecan, and oxaliplatin combinations in patients with previously untreated metastatic colorectal cancer.  J Clin Oncol2004; 22:23-30.
  30. Tournigand C, Andre T, Achille E, et al: FOLFIRI followed by FOLFOX6 or the reverse sequence in advanced colorectal cancer: a randomized GERCOR study.  J Clin Oncol2004; 22:229-237.
  31. Hurwitz H, Fehrenbacher L, Novotny W, et al: Bevacizumab plus irinotecan, fluorouracil, and leucovorin for metastatic colorectal cancer.  N Engl J Med2004; 350:2335-2342.
  32. Allen-Mersh T, Earlam S, Fordy C, et al: Quality of life and survival with continuous hepatic artery floxuridine infusion for colorectal liver metastases.  Lancet1994; 344:1255-1260.
  33. Glimelus B, Hoffman K, Olafsdottir M, et al: Quality of life during cytostatic therapy for advanced symptomatic colorectal carcinoma: a randomized comparison of two regimens.  Eur J Cancer Clin Oncol1989; 25:829-835.
  34. Scheithauer W, Rosen H, Kornek GV, et al: Randomized comparison of combination chemotherapy plus supportive care versus supportive care alone in patients with metastatic colorectal cancer.  BMJ1993; 306:752-755.
  35. Breedis C, Young C: The blood supply of neoplasms in the liver.  Am J Pathol1954; 30:969-977.
  36. Ensminger WD, Gyves JW: Hepatic arterial infusion chemotherapy for metastatic liver cancers.  Compr Ther1984; 10:25-34.
  37. Ensminger WD, Rosowsky A, Raso V, et al: A clinical-pharmacological evaluation of hepatic arterial infusions of 5-fluoro-2′-deoxyuridine and 5-fluorouracil.  Cancer Res1978; 38:3784-3792.
  38. Van Groeningen C, Van der Vish W, Giaccone G, et al: Phase I clinical and pharmacokinetic study of 5 day CPT-11 hepatic arterial infusion (HAI) chemotherapy.  Proc ASCO1997; 16:A768.
  39. Collins JM: Pharmacologic rationale for regional drug delivery.  J Clin Oncol1984; 2:498-504.
  40. Weiss L, Grandmann E, Torhost J, et al: Hematogenous metastatic patterns in colonic carcinoma: an analysis of 1541 necropsies.  J Pathol1986; 150:195-203.
  41. Weiss L: Metastatic inefficiency and regional therapy for liver metastases from colorectal carcinoma.  Regul Cancer Treat1989; 2:77-81.
  42. Yasuda S, Noto T, Ikeda M, et al: Hepatic arterial infusion chemotherapy using implantable reservoir in colorectal liver metastasis.  GanTo Kagaku Ryoho1990; 8:1815-1819.
  43. Kemeny N, Niedzwiecki D, Shurgot B, Oderman P: Prognostic variables in patients with hepatic metastases from colorectal cancer. Importance of medical assessment of liver involvement.  Cancer1989; 63:742-747.
  44. Kemeny N, Braun Jr DW: Prognostic factors in advanced colorectal carcinoma. Importance of lactic dehydrogenase level, performance status, and white blood cell count.  Am J Med1983; 74:786-794.
  45. Chang AE, Schneider PD, Sugarbaker PH: A prospective randomized trial of regional versus systemic continuous 5-fluorodeoxyuridine chemotherapy in the treatment of colorectal liver metastases.  Ann Surg1987; 206:685-693.
  46. Allen-Mersh TG, Earlam S, Fordy C, et al: Quality of life and survival with continuous hepatic-artery floxuridine infusion for colorectal liver metastases.  Lancet1994; 344:1255-1260.
  47. Lorenz M, Muller HH: Randomized, multicenter trial of fluorouracil plus leucovorin administered either via hepatic arterial or intravenous infusion versus fluorodeoxyuridine administered via hepatic arterial infusion in patients with nonresectable liver metastases from colorectal carcinoma.  J Clin Oncol2000; 18:243-254.
  48. Kerr DJ, McCardle CS, Ledermann J, et al: Intrahepatic arterial versus intravenous fluorouracil and folinic acid for colorectal cancer liver metastases: a multicentre randomised trial.  Lancet2003; 361:368-373.
  49. Kemeny NE, Niedzwiecki D, Hollis DR, et al: Hepatic arterial infusion versus systemic therapy for hepatic metastases from colorectal cancer: a randomized trial of efficacy, quality of life, and molecular markers (CALGB 9481).  J Clin Oncol2006; 24:1395-1403.
  50. Kemeny N, Seiter K, Niedzwiecki D, et al: A randomized trial of intrahepatic infusion of fluorodeoxyuridine with dexamethasone versus fluorodeoxyuridine alone in the treatment of metastatic colorectal cancer.  Cancer1992; 69:327-334.
  51. Meta-Analysis GiC: Reappraisal of HAI in the treatment of nonresectable liver metastases from colorectal carcinoma.  J Natl Cancer Inst1996; 88:252-258.
  52. Hrushesky W, Von Roemelling R, Lanning R, Rabtini J: Circadian-shaped infusions of floxuridine for progressive metastatic renal cell carcinoma.  J Clin Oncol1990; 8:1504-1513.
  53. Stagg R, Venook A, Chase J, et al: Alternating hepatic intraarterial floxuridine and fluorouracil: a less toxic regimen for treatment of liver metastases from colorectal cancer.  J Natl Cancer Inst1991; 83:423-428.
  54. Davidson B, Izzo F, Chase J, et al: Alternating floxuridine and 5-fluorouracil hepatic arterial chemotherapy for colorectal liver metastases minimizes biliary toxicity.  Am J Surg1996; 172:244-247.
  55. Patt Y, Charnsangavej C, Yoffe B, et al: Hepatic arterial infusion of floxuridine, leucovorin, doxorubicin, and cisplatin for hepatocellular carcinoma: effects of hepatitis B and C viral infection on drug toxicity and patient survival.  J Clin Oncol1994; 12:1204-1211.
  56. Warren H, Anderson J, O'Gorman P, et al: A phase II study of regional 5-fluorouracil infusion with intravenous folinic acid for colorectal liver metastases.  Br J Cancer1994; 70:677-680.
  57. Metzger U, Weder W, Rothlin M, Largiader F: Phase II study of intraarterial fluorouracil and mitomycin-C for liver metastases of colorectal cancer.  Recent Results Cancer Res1991; 121:198-204.
  58. Schlag P, Hohenberger P, Holting T, et al: Hepatic arterial infusion (HAI) chemotherapy for liver metastases of colorectal cancer using 5-FU.  Eur J Surg Oncol1990; 16:99-104.
  59. Kemeny N, Cohen A, Steiner K, et al: Randomized trial of hepatic arterial FUDR, mitomycin and BCNU versus FUDR alone: effective salvage therapy for liver metastases of colorectal cancer.  J Clin Oncol1993; 11:330-335.
  60. Kemeny N, Eid A, Stockman J, et al: Hepatic arterial infusion of floxuridine and dexamethasone plus high-dose mitomycin C for patients with unresectable hepatic metastases from colorectal carcinoma.  J Surg Oncol2005; 91:97-101.
  61. Kemeny N, Cohen A, Bertino JR, et al: Continuous intrahepatic infusion of floxuridine and leucovorin through an implantable pump for the treatment of hepatic metastases from colorectal carcinoma.  Cancer1990; 65:2446-2450.
  62. Kemeny N, Conti JA, Cohen A, et al: Phase II study of hepatic arterial floxuridine, leucovorin, and dexamethasone for unresectable liver metastases from colorectal carcinoma.  J Clin Oncol1994; 12:2288-2295.
  63. Safi F, Bittner R, Roscher R, et al: Regional chemotherapy for hepatic metastases of colorectal carcinoma (continuous intraarterial versus continuous intraarterial/intravenous therapy).  Cancer1989; 64:379-387.
  64. Wanebo H, Levy A, Vezeridis M, Cummings F: Hepatic artery infusion (HAI) of 5-fluorodeoxyuridine (FUdR) and dexamethasone (Dex) with low dose leucovorin (LV) (10 days) alternating with continuous intravenous infusion (CI) of 5-fluorouracil (5FU) with simultaneous low dose leucovorin (14 day) of unresectable metastases of colorectal cancer (CRC) (meeting abstract 537).  Proc Am Soc Clin Oncol1996;15.
  65. Allen-Mersh TG, Kemeny N, Niedzwiecki D, et al: Significance of a fall in serum CEA concentration in patients treated with cytotoxic chemotherapy for disseminated colorectal cancer.  Gut1987; 28:1625-1629.
  66. O'Connell M, Nagorney D, Bernath A, et al: Sequential intrahepatic fluorodeoxyuridine and systemic fluorouracil plus leucovorin for the treatment of metastatic colorectal cancer confined to the liver.  J Clin Oncol1998; 16:2528-2533.
  67. Kemeny N, Gonen M, Sullivan D, et al: Phase I study of hepatic arterial infusion of floxuridine and dexamethasone with systemic irinotecan for unresectable hepatic metastases from colorectal cancer.  J Clin Oncol2001; 19:2687-2695.
  68. Kemeny N, Jarnagin W, Paty P, et al: Phase I trial of systemic oxaliplatin combination chemotherapy with hepatic arterial infusion in patients with unresectable liver metastases from colorectal cancer.  J Clin Oncol2005; 23:4888-4896.
  69. Gluck W, Akwari O, Kelvin F, et al: A reversible enteropathy complicating continuous hepatic artery infusion chemotherapy with 5-fluoro 2-deoxyuridine.  Cancer1985; 56:2424-2427.
  70. Hohn D, Stagg R, Price D, et al: Avoidance of gastroduodenal toxicity in patients receiving hepatic arterial 5-fluoro-2-deoxyuridine.  J Clin Oncol1985; 3:1257-1260.
  71. Northover J, Terblance J: A new look at the arterial supply of the bile duct in man and its surgical implications.  Br J Surg1979; 66:379-384.
  72. Kemeny M, Battifora H, Flayney D, et al: Sclerosing cholangitis after continuous hepatic artery infusion of FUDR.  Ann Surg1985; 202:176-181.
  73. Brown K, Kemeny N, Berger M, et al: Obstructive jaundice in patients receiving hepatic artery infusional chemotherapy: etiology, treatment implications, and complications after transhepatic biliary drainage.  J Vasc Interv Radiol1997; 8:229-234.
  74. Campbell KA, Burns RC, Sitzmann JV, et al: Regional chemotherapy devices: effect of experience and anatomy on complications.  J Clin Oncol1993; 11:822-826.
  75. Kornmann M, Link K, Lenz H, et al: Thymidylate synthase is a predictor for response and resistance in hepatic artery infusion chemotherapy.  Cancer Lett1997; 118:29-35.
  76. Belluco C, Guillem J, Kemeny N, et al: p53 nuclear protein overexpression in colorectal cancer: a dominant predictor of survival in patients with advanced hepatic metastases.  J Clin Oncol1996; 14:2696-2701.
  77. Danso M, Jarnagin WR, Muruganandham M, et al: Hepatic arterial infusion (HAI) therapy in patients with unresectable primary liver cancer: use of dynamic contrast enhanced MRI to evaluate response.  J Clin Oncol2005; 23(16 Suppl):4129.
  78. Yoshikawa M, Sunaga M, Ebara M, et al: Hepatic arterial infusion chemotherapy by using reservoir for the treatment of advanced HCC.  Liver Int2005; 25:1348.
  79. Giaccheti S, Zidani R, Perpoint B, et al: Phase III trial of l5-fluorouracil (5-FU), folinic acid (FA), with or without oxaliplatin (OXA) in previously untreated patients (pts) with metastatic colorectal cancer (MCC) (abstract).  Proc Am Soc Clin Oncol1997; 16:229a.
  80. Bismuth H, Adam R, Levi F: Resection of nonresectable liver metastases from colorectal cancer after neoadjuvant chemotherapy.  Farabos Ann Surg1996; 224:509-520.
  81. Alberts SR, Horvath WL, Sternfeld WC, et al: Oxaliplatin, fluorouracil, and leucovorin for patients with unresectable liver-only metastases from colorectal cancer: a North Central Cancer Treatment Group phase II study.  J Clin Oncol2005; 23:9243-9249.
  82. Pozzo C, Basso M, Quirino M, et al: Long-term followup of colorectal cancer (CRC) patients treated with neoadjuvant chemotherapy with irinotecan and fluorouracil plus folinic acid (5-FU/FA) for unresectable liver metastases. In: 2006 ASCO Annual Meeting. Atlanta, Georgia, USA: American Society of Clinical Oncology, 2006, 3576.
  83. Rougier P, Raoul J-L, Van Laethem J-L, et al: Cetuximab + FOLFIRI as first-line treatment for metastatic colorectal CA. In: 2004 ASCO Annual Meeting. New Orleans, Louisiana, USA, 2004, 3513.
  84. Falcone A, Masi G, Brunetti I, et al: The triplet combination of irinotecan, oxaliplatin and 5FU/LV (FOLFOXIRI) vs the doublet of irinotecan and 5FU/LV (FOLFIRI) as first-line treatment of metastatic colorectal cancer (MCRC): results of a randomized phase III trial by the Gruppo Oncologico Nord Ovest (G.O.N.O.). In: 2006 ASCO Annual Meeting. Atlanta, Georgia, USA: American Society of Clinical Oncology; 2006.
  85. Adam R, Pascal G, Castaing D, et al: Tumor progression while on chemotherapy: a contraindication to liver resection for multiple colorectal metastases?.  Ann Surg2004; 240:1052-1061.
  86. van Haarst JM, Baas P, Manegold C, et al: Multicentre phase II study of gemcitabine and cisplatin in malignant pleural mesothelioma.  Br J Cancer2002; 86:342-345.
  87. Steele JP, Shmash J, Evans MT, et al: Phase II study of vinorelbine in patients with malignant pleural mesothelioma.  J Clin Oncol2000; 18:3912-3917.
  88. Huitzil-Melendez F, Capanu M, Haviland D, Kemeny NE: Evaluation of the impact of systemic (SYS) neoadjuvant chemotherapy (neoadj) in patients (pts) with resectable liver metastasis (mets) from colorectal carcinoma (CRC) treated with adjuvant hepatic arterial infusion (HAI) and SYS chemotherapy. In: 2007 Gastrointestinal Cancers Symposium. Orlando, Florida, USA: American Society of Clinical Oncology, 2007, 14503.
  89. Benoist S, Brouquet A, Penna C, et al: Complete response of colorectal liver metastases after chemotherapy: does it mean cure?.  J Clin Oncol2006; 24:3939-3945.
  90. Rubbia-Brandt L, Audard V, Sartoretti P, et al: Severe hepatic sinusoidal obstruction associated with oxaliplatin-based chemotherapy in patients with metastatic colorectal cancer.  Ann Oncol2004; 15:460-466.
  91. Kooby DA, Fong Y, Suriawinata A, et al: Impact of steatosis on perioperative outcome following hepatic resection.  J Gastrointest Surg2003; 7:1034-1044.
  92. Vauthey JN, Pawlik TM, Ribero D, et al: Chemotherapy regimen predicts steatohepatitis and an increase in 90-day mortality after surgery for hepatic colorectal metastases.  J Clin Oncol2006; 24:2065-2072.
  93. Gruenberger T, Sorbye H, Debois M, et al: Tumor response to preoperative chemotherapy (CT) with FOLFOX-4 for resectable colorectal cancer liver metastases (LM). Interim results of EORTC Intergroup randomized phase III study 40983. In: 2006 ASCO Annual Meeting. Atlanta, Georgia, USA: American Society of Clinical Oncology, 2006, 3500.
  94. Markowitz J: The hepatic artery.  Surg Gynecol Obstet1952; 95:644-646.
  95. Hunt TM, Flowerdew AD, Birch SJ, et al: Prospective randomized controlled trial of hepatic arterial embolization or infusion chemotherapy with 5-fluorouracil and degradable starch microspheres for colorectal liver metastases.  Br J Surg1990; 77:779-782.
  96. Gerard A, Buyse M, Pector JC, et al: Hepatic artery ligation with and without portal infusion of 5-FU. A randomized study in patients with unresectable liver metastases from colorectal carcinoma. The E.O.R.T.C. Gastrointestinal Cancer Cooperative Group (G.I. Group).  Eur J Surg Oncol1991; 17:289-294.
  97. Martensson H, Nobin A, Bengmark S, et al: Embolization of the liver in the management of metastatic carcinoid tumors.  J Surg Oncol1984; 27:152-158.
  98. Ajani J, Carrasco CH, Charnsanpavej C, et al: Islet cell tumors metastatic to the liver: effective palliation by sequential hepatic artery embolization.  Ann Intern Med1988; 108:340-344.
  99. Carrasco C, Charnsanparej C, Ajani J, et al: The carcinoid syndrome: palliation by hepatic artery embolization.  Am J Roentgenol1986; 147:149-154.
  100. Marlink R, Lokich J, Robins J, Clouse M: Hepatic arterial embolization for metastatic hormone-secreting tumors.  Cancer1990; 65:2227-2232.
  101. Carr BI, Zajko A, Bron K, et al: Phase II study of Spherex (degradable starch microspheres) injected into the hepatic artery in conjunction with doxorubicin and cisplatin in the treatment of advanced-stage hepatocellular carcinoma: interim analysis.  Semin Oncol1997; 24(Suppl 6):97-99.
  102. Bruix J, Llovet J, Castells A, et al: Transarterial embolization versus symkptomatic treatment in patients with advanced hepatocellular carcinoma: results of a randomized controlled trial in a single institution.  Hepatology1998; 27:1578-1583.
  103. Martin J, Moertel C, Adson M, et al: Surgical treatment of functioning metastatic carcinoid tumors.  Arch Surg1983; 118:537-541.
  104. Moertel C: Treatment of the carcinoid tumor and the malignant carcinoid syndrome.  J Clin Oncol1983; 1:727-740.
  105. Melia W, Nunnerly H, Johnson P, et al: Use of arterial devascularization and cytotoxic drugs in 30 patients with the carcinoid syndrome.  Br J Cancer1982; 46:331-339.
  106. Kvols L, Buck M, Moertel C, et al: Treatment of metastatic islet cell carcinoma with a somatostatin analogue (SMS 201–995).  Ann Intern Med1987; 107:162-168.
  107. Daniels J, Daniels A, Quinn M, et al: Phase I trial with cisplatin or mitomycin hepatic chemoembolization (CE) with Angiostat collagen for embolization (CFE) in patients with colorectal cancer.  Proc ASCO1988; 7:385.
  108. Venook A, Stagg R, Lewis B, et al: Chemoembolization for hepatocellular carcinoma.  J Clin Oncol1990; 8:1108-1114.
  109. Raoul J, Bourquet P, Bretagne J, et al: Hepatic artery injection of I-131 labeled Lipidol: Part I. Biodistribution study in patients with hepatocellular carcinoma and liver metastases.  Radiology1988; 168:541-545.
  110. Popov I, Lavrnic S, Jelic S, et al: Chemoembolization for liver metastases from colorectal carcinoma: risk or a benefit.  Neoplasma2002; 49:43-48.
  111. Bavisotto LM, Patel NH, Althaus SJ, et al: Hepatic transcatheter arterial chemoembolization alternating with systemic protracted continuous infusion 5-fluorouracil for gastrointestinal malignancies metastatic to liver: a phase II trial of the Puget Sound Oncology Consortium (PSOC 1104).  Clin Cancer Res1999; 5:95-109.
  112. McVie J, Hoefnagel C, Burger J: Tumor targeting of cytostatic compounds using intraarterial biodegradable microspheres.  Ann Rep Netherlands Cancer Inst1985;83-84.
  113. Sigurdson E, Ridge J, Daly J: Intraarterial infusion of doxorubicin with degradable starch microspheres.  Arch Surg1986; 121:1277-1281.
  114. Ball A: Regional chemotherapy for colorectal hepatic metastases using degradable starch micro spheres.  Acta Oncol1991; 30:301.
  115. Andrews J, Walker S, Ackermann R, et al: Hepatic radioembolization with Yttrium-90 containing glass microspheres: preliminary results and clinical followup.  J Nucl Med1994; 35:1637-1644.
  116. Bretagne JF, Raoul JL, Bourguet P, et al: Hepatic artery injection of I-131-labeled lipiodol. Part II. Preliminary results of therapeutic use in patients with hepatocellular carcinoma and liver metastases.  Radiology1988; 168:547-550.
  117. Pelletier G, Roche A, Ink O, et al: A randomized trial of hepatic arterial chemoembolization in patients with unresectable hepatocellular carcinoma.  J Hepatol1990; 11:181-184.
  118. Groupe d'Etude et de Traitement du Carcinome Hépatocellulaire : A comparison of lipiodol chemoembolization and conservative treatment for unresectable hepatocellular carcinoma.  N Engl J Med1995; 332:1256-1261.
  119. Pelletier G, Ducreux M, Gay F, et al: Treatment of unresectable hepatocellular carcinoma with lipiodol chemoembolization: a multicenter randomized trial. Groupe CHC.  J Hepatol1998; 29:129-134.
  120. Llovet JM, Real MI, Montana X, et al: Arterial embolisation or chemoembolisation versus symptomatic treatment in patients with unresectable hepatocellular carcinoma: a randomised controlled trial.  Lancet2002; 359:1734-1739.
  121. Ravikumar T, Kane R, Cady B, et al: A 5-year study of cryosurgery in the treatment of liver tumors.  Arch Surg1991; 126:1520-1524.
  122. Onik G, Rubinsky B, Zemel R, et al: Ultrasound-guided hepatic cryosurgery in the treatment of metastatic colon carcinoma.  Cancer1991; 67:901-907.
  123. Weaver ML, Atkinson D, Zemel R: Hepatic cryosurgery in treating colorectal metastases.  Cancer1995; 76:210-214.
  124. Morris D, Ross W: Australian experience of cryoablation of liver tumors: metastases.  Surg Oncol Clin North Am1996; 5:391-397.
  125. Steele Jr G, Ravikumar T, Benotti P: New surgical treatments for recurrent colorectal cancer (review).  Cancer1990; 65:723-730.
  126. Adam R, Akpinar E, Johann M, et al: Place of cryosurgery in the treatment of malignant liver tumors.  Ann Surg1997; 225:39.38
  127. Preketes A, Caplehorn J, King J, et al: Effect of hepatic artery chemotherapy on survival of patients with hepatic metastases from colorectal carcinoma treated with cryotherapy.  World J Surg1995; 19:768-771.
  128. Litvak DA, Wood TF, Tsioulias GJ, et al: Systemic irinotecan and regional floxuridine after hepatic cytoreduction in 185 patients with unresectable colorectal cancer metastases.  Ann Surg Oncol2002; 9:148-155.
  129. Welling R, Lamping K: Cryoprobe as a “handle” for resection of metastatic liver tumors.  J Surg Oncol1990; 45:227-228.
  130. Ravikumar TS, Steele GJ, Kane R, et al: Experimental and clinical observations on hepatic cryosurgery for colorectal metastases.  Cancer Res1991; 51:6233-6237.
  131. Tandan V, Harmantas A, Gallinger S: Long-term survival after hepatic cryosurgery versus surgical resection for metastatic colorectal carcinoma: a critical review of the literature.  Can J Surg1997; 40:175-181.
  132. Solbiati L, Ierace T, Tonolini M, et al: Radiofrequency thermal ablation of hepatic metastases.  Eur J Ultrasound2001; 13:149-158.
  133. Bilchik AJ, Wood TF, Allegra D, et al: Cryosurgical ablation and radiofrequency ablation for unresectable hepatic malignant neoplasms: a proposed algorithm.  Arch Surg2000; 135:657-662.
  134. Curley SA, Izzo F, Delrio P, et al: Radiofrequency ablation of unresectable primary and metastatic hepatic malignancies: results in 123 patients.  Ann Surg1999; 230:1-8.
  135. Sugiyama A, Katayama M, Matsuda T, et al: Hepatic arterial infusion chemotherapy combined with hyperthermia for metastatic liver tumors of colorectal cancer.  Semin Oncol1997; 24(Suppl 6):135-138.
  136. Sata M, Watanabe Y, Ueda S, et al: Microwave coagulation therapy for hepatocellular carcinoma.  Gastroenterology1996; 110:1507-1514.
  137. Lencioni R, Crocetti L, Cioni D, et al: Percutaneous radiofrequency ablation of hepatic colorectal metastases: technique, indications, results, and new promises.  Invest Radiol2004; 39:689-697.
  138. Cheung L, van Sonnenberg E, Morrison P, et al: Radiofrequency ablation therapy.  Contemp Diagn Radiol2005; 28:1-5.
  139. Elias D, De Baere T, Smayra T, et al: Percutaneous radiofrequency thermoablation as an alternative to surgery for treatment of liver tumour recurrence after hepatectomy.  Br J Surg2002; 89:752-756.
  140. Livraghi T, Solbiati L, Meloni F, et al: Percutaneous radiofrequency ablation of liver metastases in potential candidates for resection: the “test-of-time approach.”.  Cancer2003; 97:3027-3035.
  141. Berber E, Pelley R, Siperstein AE: Predictors of survival after radiofrequency thermal ablation of colorectal cancer metastases to the liver: a prospective study.  J Clin Oncol2005; 23:1358-1364.
  142. Solbiati L, Livraghi T, Goldberg SN, et al: Percutaneous radiofrequency ablation of hepatic metastases from colorectal cancer: long-term results in 117 patients.  Radiology2001; 221:159-166.
  143. Scaife CL, Curley SA, Izzo F, et al: Feasibility of adjuvant hepatic arterial infusion of chemotherapy after radiofrequency ablation with or without resection in patients with hepatic metastases from colorectal cancer.  Ann Surg Oncol2003; 10:348-354.
  144. Kainuma O, Asano T, Aoyama H, et al: Combined therapy with radiofrequency thermal ablation and intraarterial infusion chemotherapy for hepatic metastases from colorectal cancer.  Hepatogastroenterology1999; 46:1071-1077.
  145. Martin RC, Scoggins CR, McMasters KM: A phase II study of radiofrequency ablation of unresectable metastatic colorectal cancer with hepatic arterial infusion pump chemotherapy.  J Surg Oncol2006; 93:387-393.
  146. Chow DH, Sinn LH, Ng KK, et al: Radiofrequency ablation for hepatocellular carcinoma and metastatic liver tumors: a comparative study.  J Surg Oncol2006; 94:565-571.
  147. Henn AR, Levine EA, McNulty W, Zagoria RJ: Percutaneous radiofrequency ablation of hepatic metastases for symptomatic relief of neuroendocrine syndromes.  Am J Roentgenol2003; 181:1005-1010.
  148. Wettstein M, Vogt C, Cohnen M, et al: Serotonin release during percutaneous radiofrequency ablation in a patient with symptomatic liver metastases of a neuroendocrine tumor.  Hepatogastroenterology2004; 51:830-832.
  149. Livraghi T, Goldberg SN, Solbiati L, et al: Percutaneous radiofrequency ablation of liver metastases from breast cancer: initial experience in 24 patients.  Radiology2001; 220:145-149.
  150. de Baere T, Risse O, Kuoch V, et al: Adverse events during radiofrequency treatment of 582 hepatic tumors.  Am J Roentgenol2003; 181:695-700.
  151. Wood BJ, Ramkaransingh JR, Fojo T, et al: Percutaneous tumor ablation with radiofrequency.  Cancer2002; 94:443-451.
  152. Livraghi T, Baietta E, Matricardi L, et al: Fine needle percutaneous intratumoral chemotherapy under ultrasound guidance: a feasibility study.  Tumori1986; 72:81-87.
  153. Suzuki M, Suzuki H, Yamamoto T, et al: Indi-cation of chemoembolization therapy without gelatin sponge for hepatocellular carcinoma.  Semin Oncol1997; 24(Suppl 6):56-110.
  154. Shiina S, Imamura M, Obi S, et al: Percutaneous ethanol injection therapy for small hepatocellular carcinoma.  Gan to Kagaku Ryoho1996; 23:835-839.
  155. Yamamoto K, Masuzawa M, Kato M, et al: Evaluation of comobined therapy with chemoembolization and ethanol injection for advanced hepatocellular carcinoma.  Semin Oncol1997; 24(Suppl 6):50-55.
  156. Marinelli A, de Brauw LM, Beerman H, et al: Isolated liver perfusion with mitomycin C in the treatment of colorectal cancer metastases confined to the liver.  Jpn J Clin Oncol1996; 26:341-350.
  157. Marinelli A, Vahrmeijer AL, van de Velde CJ: Phase I/II studies of isolated hepatic perfusion with mitomycin C or melphalan in patients with colorectal cancer hepatic metastases.  Recent Results Cancer Res1998; 147:83-94.
  158. Alexander H, Bartlett D, Libutti S, et al: Isolated hepatic perfusion with tumor necrosis factor and melphalan for unresectable cancers confined to the liver.  J Clin Oncol1998; 16:1479-1489.
  159. Venook A, Bergsland E, Ring E, et al: Gene therapy of colorectal liver metastases using a recombinant adenovirus encoding wt p53 (SCH 58500) via hepatic artery infusion: a phase I study [abstract].  Proc Am Soc Clin Oncol1998; 17:431a.
  160. Trinh Q, Austin E, Murray D, et al: Enzyme/prodrug gene therapy: comparison of cytosine deaminase/5-fluorocytosine versus thymidine kinase/ganciclovir enzyme/prodrug systems in a human colorectal carcinoma cell line.  Cancer Res1995; 55:4808-4812.
  161. Rubin J, Galanis E, Pitot H, et al: Phase I study of immunotherapy of hepatic metastases of colorectal carcinoma by direct gene transfer of an allogeneic histocompatibility antigen, HLA-B7.  Gene Ther1997; 4:419-425.
  162. Russell AH, Clyde C, Wasserman TH, et al: Accelerated hyperfractionated hepatic irradiation in the management of patients with liver metastases: results of the RTOG dose escalating protocol.  Int J Radiat Oncol Biol Phys1993; 27:117-123.
  163. Borgelt BB, Gelber R, Brady LW, et al: The palliation of hepatic metastases: results of the Radiation Therapy Oncology Group pilot study.  Int J Radiat Oncol Biol Phys1981; 7:587-591.
  164. Phillips R, Karnofsky D, Hamilton L, Nickson J: Roentgen therapy of hepatic metastases.  Am J Roentgenol1954; 71:826-834.
  165. Prasad B, Lee M, Hendrickson F: Irradiation of hepatic metastases.  Int J Radiat Oncol Biol Phys1977; 2:129-132.
  166. Leibel SA, Pajak TF, Massullo V, et al: A comparison of misonidazole sensitized radiation therapy to radiation therapy alone for the palliation of hepatic metastases: results of a Radiation Therapy Oncology Group randomized prospective trial.  Int J Radiat Oncol Biol Phys1987; 13:1057-1064.
  167. Bydder S, Spry N, Christie D, et al: A prospective trial of short-fractionation radiotherapy for the palliation of liver metastases.  Australas Radiol2003; 47:284-288.
  168. Heimburger D, Shewach D, Lawrence T: The effect of fluordeoxyuridine on sublethal damage repair in human colon cancer cells.  Int J Radiat Oncol Biol Phys1991; 21:983-987.
  169. Bruso C, Shewach D, Lawrence T: Fluorodeoxyuridine-induced radiosensitization and inhibition of DNA soluble strand break repair in human colon cancer cells.  Int J Radiat Oncol Biol Phys1990; 19:1411-1417.
  170. Byfield JE, Calabro-Jones P, Klisak I, Kulhanian F: Pharmacologic requirements for obtaining sensitization of human tumor cells in vitro to combined 5-fluorouracil or ftorafur and X rays.  Int J Radiat Oncol Biol Phys1982; 8:1923-1933.
  171. Witte R, Cnaan A, Mansour E, et al: Comparison of 5-fluoruracil alone, 5-fluorouracil with levamisole, and 5-fluorouavin with hepatic irradiation in the treatment of patients with residual, nonmeasureable intraabdominal metastases after undergoing resection for colorectal carcinoma.  Cancer2001; 91:1020-1028.
  172. Ten Haken RK, Lawrence TS, McShan DL, et al: Technical considerations in the use of 3-D beam arrangements in the abdomen.  Radiother Oncol1991; 22:19-28.
  173. Mohiuddin M, Chen E, Ahmad N: Combined liver radiation and chemotherapy for palliation of hepatic metastases from colorectal cancer.  J Clin Oncol1996; 14:722-728.
  174. Ben-Josef E, Normolle D, Ensminger WD, et al: Phase II trial of high-dose conformal radiation therapy with concurrent hepatic artery floxuridine for unresectable intrahepatic malignancies.  J Clin Oncol2005; 23:8739-8747.
  175. Schefter TE, Kavanagh BD, Timmerman RD, et al: A phase I trial of stereotactic body radiation therapy (SBRT) for liver metastases.  Int J Radiat Oncol Biol Phys2005; 62:1371-1378.
  176. Robertson JM, Lawrence TS, Walker S, et al: The treatment of colorectal liver metastases with conformal radiation therapy and regional chemotherapy.  Int J Radiat Oncol Biol Phys1995; 32:445-450.
  177. Dawson LA, McGinn CJ, Normolle D, et al: Escalated focal liver radiation and concurrent hepatic artery fluorodeoxyuridine for unresectable intrahepatic malignancies.  J Clin Oncol2000; 18:2210-2218.
  178. Blomgren H, Lax I, Naslund I, Svanstrom R: Stereotactic high dose fraction radiation therapy of extracranial tumors using an accelerator. Clinical experience of the first thirty-one patients.  Acta Oncol1995; 34:861-870.
  179. Wulf J, Hadinger U, Oppitz U, et al: Stereotactic boost irradiation for targets in the abdomen and pelvis.  Radiother Oncol2004; 70:31-36.
  180. Herfarth KK, Debus J, Wannenmacher M: Stereotactic radiation therapy of liver metastases: update of the initial phase-I/II trial.  Front Radiat Ther Oncol2004; 38:100-105.
  181. Fuss M, Thomas Jr CR: Stereotactic body radiation therapy: an ablative treatment option for primary and secondary liver tumors.  Ann Surg Oncol2004; 11:130-138.
  182. Katz A, Carey-Sampson M, Muhs A, et al: Hypofractionated stereotactic body radiation therapy (SBRT) for limited hepatic metastases.  Int J Radiat Oncol Biol Phys2007; 67:793.789
  183. Mendez Romero A, Wunderink W, Hussain SM, et al: Stereotactic body radiation therapy for primary and metastatic liver tumors: a single institution phase I-II study.  Acta Oncol2006; 45:831-837.
  184. Kavanagh BD, Schefter TE, Cardenes HR, et al: Interim analysis of a prospective phase I/II trial of SBRT for liver metastases.  Acta Oncol2006; 45:848-855.
  185. Blomgren J, Lax I, Goranson H, et al: Radiosurgery for tumors in the body: clinical experience using a new method.  J Radiosurg1998; 1:63-74.
  186. Gunven P, Blomgren H, Lax I: Radiosurgery for recurring liver metastases after hepatectomy2003; 50:1201-1204.
  187. Herfarth KK, Debus J, Lohr F, et al: Stereotactic single-dose radiation therapy of liver tumors: results of a phase I/II trial.  J Clin Oncol2001; 19:164-170.
  188. Wulf J, Hadinger U, Oppitz U, et al: Stereotactic radiotherapy of targets in the lung and liver.  Strahlenther Onkol2001; 177:645-655.
  189. Dawson LA, Eccles C, Craig T: Individualized image guided iso-NTCP based liver cancer SBRT.  Acta Oncol2006; 45:856-864.
  190. Masters A, Steger AC, Bown SG: Role of interstitial therapy in the treatment of liver cancer.  Br J Surg1991; 78:518-523.
  191. Ricke J, Wust P, Wieners G, et al: Liver malignancies: CT-guided interstitial brachytherapy in patients with unfavorable lesions for thermal ablation.  J Vasc Interv Radiol2004; 15:1279-1286.
  192. Donath D, Nori D, Turnbull A, et al: Brachytherapy in the treatment of solitary colorectal metastases to the liver.  J Surg Oncol1990; 44:55-61.
  193. Thomas DS, Nauta RJ, Rodgers JE, et al: Intraoperative high-dose rate interstitial irradiation of hepatic metastases from colorectal carcinoma. Results of a phase I–II trial.  Cancer1993; 71:1977-1981.
  194. Armstrong J, Anderson L, Harrison L: Treatment of liver metastses from colorectal cancer with radioactive implants.  Cancer1994;1800-1804.
  195. Welsh JS, Kennedy AS, Thomadsen B: Selective Internal Radiation Therapy (SIRT) for liver metastases secondary to colorectal adenocarcinoma.  Int J Radiat Oncol Biol Phys2006; 66(2 Suppl):S62-S73.
  196. Gray B, Hazel GV, Hope M, et al: Randomised trial of SIR-Spheres plus chemotherapy vs. chemotherapy alone for treating patients with liver metastases from primary large bowel cancer.  Ann Oncol2001; 12:1711-1720.
  197. Stubbs RS, Cannan RJ, Mitchell AW: Selective internal radiation therapy (SIRT) with 90Yttrium microspheres for extensive colorectal liver metastases.  Hepatogastroenterology2001; 48:333-337.
  198. Mancini R, Carpanese L, Sciuto R, et al: A multicentric phase II clinical trial on intraarterial hepatic radiotherapy with 90 yttrium SIR-spheres in unresectable, colorectal liver metastases refractory to i.v. chemotherapy: preliminary results on toxicity and response rates.  In Vivo2006; 60:711-714.
  199. Stubbs RS, O'Brien I, Correia MM: Selective internal radiation therapy with 90Y microspheres for colorectal liver metastases: single-centre experience with 100 patients (see comment).  Aust N Z J Surg2006; 76:696-703.
  200. Kennedy AS, Coldwell D, Nutting C, et al: Resin 90Y-microsphere brachytherapy for unresectable colorectal liver metastases: modern USA experience.  Int J Radiat Oncol Biol Phys2006; 65:412-425.
  201. Tian JH, Xu BX, Zhang JM, et al: Ultrasound-guided internal radiotherapy using yttrium-90-glass microspheres for liver malignancies (see comment).  J Nucl Med1996; 37:958-963.
  202. Kennedy A, Nutting C, Coldwell D, et al: Pathological response and microdosimetry of 90 Yttrium microspheres in man: review of four explanted whole livers.  Int J Radiat Oncol Biol Phys2004; 60:1552-1563.
  203. Wharton JT, Delclos L, Gallager S, Smith JP: Radiation hepatitis induced by abdominal irradiation with the cobalt 60 moving strip technique.  Am J Roentgenol Radium Ther Nucl Med1973; 117:73-80.
  204. Perez C, Korba A, Zivnuska F, et al: 60Co moving strip technique in the management of the ovary: analysis of tumor control and morbidity.  Int J Radiat Oncol Biol Phys1978; 4:279-388.
  205. Lawrence TS, Robertson JM, Anscher MS, et al: Hepatic toxicity resulting from cancer treatment.  Int J Radiat Oncol Biol Phys1995; 31:1237-1248.
  206. Haddad E, Le Bourgeois JP, Kuentz M, Lobo P: Liver complications in lymphomas treated with a combination of chemotherapy and radiotherapy: preliminary results.  Int J Radiat Oncol Biol Phys1983; 9:1313-1319.
  207. Schacter L, Crum E, Spitzer T, et al: Fatal radiation hepatitis: a case report and review of the literature.  Gynecol Oncol1986; 24:373-380.
  208. Ingold J, Reed G, Kaplan H: Radiation hepatitis.  Am J Roentgenol1965; 93:200-208.
  209. Kutcher GJ, Burman C: Calculation of complication probability factors for non-uniform normal tissue irradiation: the effective volume method (see comment).  Int J Radiat Oncol Biol Phys1989; 16:1623-1630.
  210. Lyman J: Complication probability as assessed from dose-volume histograms.  Radiat Res1985; 8(Suppl):S13-S19.
  211. Niemierko A, Goitein M: Modeling of normal tissue response to radiation: the critical volume model.  Int J Radiat Oncol Biol Phys1993; 25:135-145.
  212. Emami B, Lyman J, Brown A, et al: Tolerance of normal tissue to therapeutic irradiation.  Int J Radiat Oncol Biol Phys1991; 21:109-122.
  213. Dawson LA, Normolle D, Balter JM, et al: Analysis of radiation-induced liver disease using the Lyman NTCP model.  Int J Radiat Oncol Biol Phys2002; 53:810-821.
  214. Bonner J, Harari P, Giralt J, et al: Radiotherapy plus cetuximab for squamous-cell carcinoma of the head and neck.  N Engl J Med2006; 354:567-578.
  215. Symon Z, Levi M, Ensminger WD, et al: Selective radioprotection of hepatocytes by systemic and portal vein infusions of amifostine in a rat liver tumor model.  Int J Radiat Oncol Biol Phys2001; 50:473-478.
  216. Scheele J, Stang R, Altendorf-Hofmann A, Paul M: Resection of colorectal liver metastases.  World J Surg1995; 19:59-71.
  217. Hohenberger P, Schlag P, Schwarz V, Herfarth C: Tumor recurrence and options for further treatment after resection of liver metastases in patients with colorectal cancer.  J Surg Oncol1990; 44:245-251.
  218. Fong Y, Cohen AM, Fortner JG, et al: Liver resection for colorectal metastases.  J Clin Oncol1997; 15:938-946.
  219. Rees M, Plant G, Bygrave S: Late results justify resection for multiple hepatic metastases from colorectal cancer.  Br J Surg1997; 84:1136-1140.
  220. Que FG, Nagorney DM: Resection of “recurrent” colorectal metastases to the liver.  Br J Surg1994; 81:255-258.
  221. Tandon R, Bunnell I, Copper R: The treatment of metastatic carcinoma of the liver by percutaneous selective hepatic artery infusion of 5-fluorouracil.  Surgery1973; 73:118-121.
  222. Ansfield FJ, Ramirez G, Davis HLJ, et al: Further clinical studies with intrahepatic arterial infusion with 5-fluorouracil.  Cancer1975; 36:2413-2417.
  223. Watkins EJ, Khazei A, Nahra K: Surgical basis for arterial infusion chemotherapy of disseminated carcinoma of the liver.  Surg Gynecol Obstet1970; 130:581-605.
  224. Cady B, Oberfield R: Regional infusion chemotherapy of hepatic metastases from carcinoma of the colon.  Am J Surg1974; 127:220-227.
  225. Smiley S, Schouten J, Chang A, et al: Intrahepatic arterial infusion with 5-FU for liver metastases of colorectal carcinoma.  Proc ASCO1981; 22:391.
  226. Denck H: Ergebnisse einer intraarteriellen intermittierenden Chemotherapie mit 5-FU bei Metastasenleber sowie inoperablen Tumoren des Gastrointestinal und Urogenitaltrakts.  Onkologie1984; 7:167-176.
  227. Schlag P, Hohenberger P, Herfath C: Resection of liver metastases in colorectal cancer—competitive analysis of treatment results in synchronous versus metachronous metastases.  Eur J Surg Oncol1990; 16:360-365.
  228. Rougier P, Laser P, Elias D, et al: Intraarterial hepatic chemotherapy (IAHC) for liver metastases (LM) from colorectal (CR) origin (abstract).  Proc ASCO1987; 6:369.
  229. Arai Y, Inaba Y, Takeuchi Y, Ariyoshi Y: Intermittent hepatic arterial infusion of high-dose 5FU on a weekly schedule for liver metastases from colorectal cancer.  Cancer Chemother Pharmacol1997; 40:526-530.
  230. Niederhuber J, Ensminger W, Gyves J, et al: Regional chemotherapy of colorectal cancer metastatic to the liver.  Cancer1984; 53:1336-1343.
  231. Balch CM, Urist MM: Intraarterial chemotherapy for colorectal liver metastases and hepatomas using a totally implantable drug infusion pump.  Recent Results Cancer Res1986; 100:123-147.
  232. Kemeny N, Daly J, Oderman P, et al: Hepatic artery pump infusion toxicity and results in patients with metastatic colorectal carcinoma.  J Clin Oncol1984; 2:595-600.
  233. Shepard K, Levin B, Karl R, et al: Therapy for metastatic colorectal cancer with hepatic artery infusion chemotherapy using a subcutaneous implanted pump.  J Clin Oncol1985; 3:161-169.
  234. Cohen A, Kaufman S, Wood W, et al: Regional hepatic chemotherapy using an implantable drug infusion pump.  Am J Surg1983; 145:529-533.
  235. Weiss G, Garnick M, Osteen R, et al: Long-term arterial infusion of 5-fluorouracil for liver metastases using an implantable infusion pump.  J Clin Oncol1983; 1:337-344.
  236. Schwartz S, Jones L, McCune C: Assessment of treatment of intrahepatic malignancies using chemotherapy via an implantable pump.  Ann Surg1985; 201:560-567.
  237. Johnson L, Wasserman P, Rivkin S: FUDR hepatic arterial infusion via an implantable pump for treatment of hepatic tumors.  Proc ASCO1983; 2:119.
  238. Kemeny N, Daly J, Reichman B, et al: Intrahepatic or systemic infusion of fluorodeoxyuridine in patients with liver metastases from colorectal carcinoma—a randomized trial.  Ann Intern Med1987; 107:459-465.
  239. Hohn D, Stagg R, Friedman M, et al: A randomized trial of continuous intravenous versus hepatic intraarterial floxuridine in patients with colorectal cancer metastatic to the liver: The Northern California Oncology Group Trial.  J Clin Oncol1989; 7:1646-1654.
  240. Niederhuber J: Arterial chemotherapy for metastatic colorectal cancer in the liver (abstract).  Second International Conference on Advances in Regional Cancer Therapy 1985,  W. Germany, Giessen, 1985.
  241. Martin JKJ, O'Connell MG, Wieland HS, et al: Intraarterial floxuridine vs systemic fluorouracil for hepatic metastases from colorectal cancer. A randomized trial.  Arch Surg1990; 125:1022-1027.
  242. Rougier P, Laplanche A, Huguier M, et al: Hepatic arterial infusion of floxuridine in patients with liver metastases from colorectal carcinoma: long-term results of a prospective randomized trial.  J Clin Oncol1992; 10:1112-1118.
  243. Cortesi E, Capussotti L, DiTora P, et al: Bolus vs continuous hepatic arterial infusion of cisplatin plus intravenous 5-fluorouracil chemotherapy for unresectable colorectal metastases.  Dis Colon Rectum1994; 37(2 Suppl):S138-S143.
  244. Sugihara K: Continuous hepatic arterial infusion of 5-fluorouracil for unresectable colorectal liver metastases: phase II study.  Surgery1995; 117:624-628.
  245. Kerr D, Ledermann J, McArdle C, et al: Phase I clinical and pharmacokinetic study of leucovorin and infusional hepatic arterial fluorouracil.  J Clin Oncol1995; 13:2968-2972.
  246. Howell J, McArdle C, Kerr D, et al: A phase I study of regional 2-weekly 5-fluorouracil infusion with intravenous folinic acid in the treatment of colorectal liver metastases.  Br J Cancer1997; 76:1390-1393.
  247. Patt Y, Hoque A, Lozano R, et al: Phase II trial of hepatic arterial infusion of fluorouracil and recombinant human interferon alfa-2b for liver metastases of colorectal cancer refractory to systemic fluorouracil and leucovorin.  J Clin Oncol1997; 15:1432-1438.
  248. Borner M, Laffer U, Ludwig C, et al: Effectiveness and low toxicity of hepatic artery infusion with fluorouracil and mitomycin for metastatic colorectal cancer confined to the liver. From the Swiss Group for Clinical and Epidemiological Cancer Research (SAKK).  Ann Oncol1990; 1:227-228.
  249. Kemeny N, Cohen A, Seiter K, et al: Randomized trial of hepatic arterial floxuridine, mitomycin, and carmustine versus floxuridine alone in previously treated patients with liver metastases from colorectal cancer.  J Clin Oncol1993; 11:330-335.
  250. Neyns B, Fontaine C, Delvaux G, et al: Efficacy of sequential hepatic arterial infusion of CPT-11 and MTX modulated 5-fluorouracil for patients with colorectal cancer metastatic to the liver after failure of systemic thymidylate-synthase inhibitor treatment. In 2002 ASCO Annual Meeting. Orlando, Florida, USA: American Society of Clinical Oncology, 2002.
  251. Cyjon A, Neuman-Levin M, Rakowsky E, et al: Liver metastases from colorectal cancer: regional intraarterial treatment following failure of systemic chemotherapy.  Br J Cancer2001; 85:504-508.
  252. Boige V, Lacombe S, De Baere T, et al: Hepatic arterial infusion oxaliplatin combined with intravenous 5-FU and folinic acid in nonresectable liver metastasis of colorectal cancer: a promising option for failures to systemic chemotherapy In 2003 ASCO Annual Meeting. Chicago, Ilinois, USA: American Society of Clinical Oncology, 2003.
  253. Ducreux M, Ychou M, Laplanche A, et al: Hepatic arterial oxaliplatin infusion plus intravenous chemotherapy in colorectal cancer with inoperable hepatic metastases: a trial of the gastrointestinal group of the Federation Nationale des Centres de Lutte Contre le Cancer.  J Clin Oncol2005; 23:4881-4887.
  254. Fraschini G, Charngangavej C, Carrasco C, et al: Percutaneous hepatic arterial infusion of cisplatin-vinblastine for refractory cancer metastatic to the liver.  Am J Clin Oncol1988; 11:34-38.
  255. Estape J, Daniels M, Vinolas N, et al: Combination chemotherapy with oral etoposide plus intravenous cyclophosphamide in liver metastases of breast cancer.  Am J Clin Oncol1990; 13:98-100.
  256. Fraschini G, Flesihman G, Charnsangavej C, et al: Continuous 5-day infusion of vinblastine for percutaneous hepatic arterial chemotherapy for metastatic breast cancer.  Cancer Treat Rep1987; 71:1001-1005.
  257. Fraschini G, Fleishman G, Yap H-Y, et al: Percutaneous hepatic arterial infusion of cisplatin for metastatic breast cancer.  Cancer Treat Rep1987; 71:313-315.
  258. Maral J, Baumer R, Curet P, et al: Intraarterial chemotherapy for liver metastases of colon and breast cancer (abstract). In Third European Conference on Clinical Oncology and Cancer Nursing, 1985, Stockholm, 1985, p 151.
  259. Arai Y, Sone Y, Inaba Y, et al: Hepatic arterial infusion chemotherapy for liver metastases from breast cancer.  Cancer Chemother Pharmacol1994; 33(Suppl):142-144.
  260. Tada A, Ogawa M, Ingaki J, et al: Arterial infusion of combination chemotherapy consisting of adriamycin and mitomycin-C for liver metastases of breast cancer.  Gan to Kagaku ryoho (Tokyo)1986; 13:70-74.
  261. Warren KW, Mountain JC, Lloyd-Jones W: Malignant tumours of the bile-ducts.  Br J Surg1972; 59:501-505.
  262. Makela JT, Kairaluoma MI: Superselective intraarterial chemotherapy with mitomycin for gallbladder cancer.  Br J Surg1993; 80:912-915.
  263. Melichar B, Cerman Jr J, Dvorak J, et al: Regional chemotherapy in biliary tract cancers—a single institution experience.  Hepatogastroenterology2002; 49:900-906.
  264. Mambrini A, Fiorentini G, Pennucci C, et al: Intraarterial hepatic chemotherapy combined with systemic infusion of 5-FU in patients with advanced biliary tract cancers. ASCO Annual Meeting Proceedings (post-meeting edition).  J Clin Oncol2004; 22:4197.
  265. Okuda K, Tanaka M, Shibata J, et al: Hepatic arterial infusion chemotherapy with continuous low dose administration of cisplatin and 5-fluorouracil for multiple recurrence of hepatocellular carcinoma after surgical treatment.  Oncology Rep1999; 6:587-591.
  266. Shildt R, Baker L, Stuckey W: Hepatic artery infusion (HAI) with 5-FUDR, adriamycin (A), and streptozocin (St) in unresectable hepatoma. A Southwest Oncology Group Study.  Proc ASCO1984; 3:150.
  267. Fazio N, Orsi F, Manzoni S, et al: Hepatic arterial infusion (HAI) using a percutaneous temporary catheter in patients with advanced primary liver carcinoma.  Proc Am Soc Clin Oncol2003; 18:331.
  268. Wellwood JM, Cady B, Oberfield RA: Treatment of primary liver cancer: response to regional chemotherapy.  Clin Oncol1979; 5:25-31.
  269. Carr B, Dvorchik I: Effects of cisplatin (DDP) intensity on hepatocellular carcinoma (HCC) responses and survival in 57 patients.  Proc Am Soc Clin Oncol1998; 17:288a.
  270. Cheng E, Watson RC, Fortner J, et al: Regional intraarterial infusion of cisplatin in primary liver cancer: a phase II trial.  Proc Am Soc Clin Oncol1982; 1:179.
  271. Urabe T, Kaneko S, Matsushita E, et al: Clinical pilot study of intrahepatic arterial chemotherapy with methotrexate, 5-fluorouracil, cisplatin and subcutaneous interferon-alpha-2b for patients with locally advanced hepatocellular carcinoma.  Oncology1998; 55:39-47.
  272. Atiq OT, Kemeny N, Niedzwiecki D, Botet J: Treatment of unresectable primary liver cancer with intrahepatic fluorodeoxyuridine and mitomycin C through an implantable pump.  Cancer1992; 69:920-924.
  273. Wein A, Riedel C, Bruckl W, et al: Neoadjuvant treatment with weekly high-dose 5-fluorouracil as 24-hour infusion, folinic acid and oxaliplatin in patients with primary resectable liver metastases of colorectal cancer.  Oncology2003; 64:131-138.
  274. Gaspar EM, Artigas V, Montserrat E, et al: Single centre study of L-OHP/5-FU/LV before liver surgery in patients with NOT optimally resectable colorectal cancer isolated liver metastases.  Proc Am Soc Clin Oncol2003; 22:1416.
  275. Quenet F, Nordlinger B, Rivoire M, et al: Resection of previously unresectable liver metastases from colorectal cancer (LMCRC) after chemotherapy (CT) with CPT-11/L-OHP/LV5FU (Folfirinox): a prospective phase II trial. In 2004 ASCO Annual Meeting, New Orleans, Louisiana, USA: American Society of Clinical Oncology, 2004, 3613.
  276. Abad A, Antón A, Massuti B, et al: Resectability of liver metastases (LM) in patients with advanced colorectal cancer (ACRC) after treatment with the combination of oxaliplatin (OXA), irinotecan (IRI) and 5FU.  J Clin Oncol2005; 23(Pt.1):3618.Final results of a phase II study. ASCO Annual Meeting Proceedings
  277. Falcone A, Masi G, Cupini S, et al: Surgical resection of metastases (mts) after biweekly chemotherapy with irinotecan, oxaliplatin and 5-fluorouracil/leucovorin (FOLFOXIRI) in initially unresectable metastatic colorectal cancer (MCRC). In 2004 ASCO Annual Meeting, New Orleans, Louisiana, USA: American Society of Clinical Oncology, 2004.
  278. Ho WM, Ma B, Mok T, et al: Liver resection after irinotecan, 5-fluorouracil, and folinic acid for patients with unresectable colorectal liver metastases: a multicenter phase II study by the Cancer Therapeutic Research Group.  Med Oncol2005; 22:303-312.
  279. Masi G, Cupini S, Marcucci L, et al: Treatment with 5-fluorouracil/folinic acid, oxaliplatin, and irinotecan enables surgical resection of metastases in patients with initially unresectable metastatic colorectal cancer.  Ann Surg Oncol2006; 13:58-65.
  280. Wein A, Riedel C, Kockerling F, et al: Impact of surgery on survival in palliative patients with metastatic colorectal cancer after first line treatment with weekly 24-hour infusion of high-dose 5-fluorouracil and folinic acid.  Ann Oncol2001; 12:1721-1727.
  281. Rougier P, Raoul JL, Van Laethem JL, et al: Cetuximab + FOLFIRI as first-line treatment for metastatic colorectal CA.  J Clin Oncol2004; 22(14 Suppl):3513.
  282. Maton P, Camilleri M, Griffin G, et al: Role of hepatic arterial embolization in the carcinoid syndrome.  Br Med J1983; 287:932-935.