Clinical Cancer Research Molecular Diagnostics in Cancer Therapeutic Development: Fulfilling the Promise of Personalized Medicine Infection and Cancer: Biology, Therapeutics, and Prevention
HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
Cancer Research Clinical Cancer Research
Cancer Epidemiology Biomarkers & Prevention Molecular Cancer Therapeutics
Molecular Cancer Research Cancer Prevention Research
Cancer Prevention Journals Portal Cancer Reviews Online
Annual Meeting Education Book Meeting Abstracts Online
This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF)
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Services
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrow reprints & permissions
Citing Articles
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Bavisotto, L. M.
Right arrow Articles by Thomas, C. R.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Bavisotto, L. M.
Right arrow Articles by Thomas, C. R., Jr.
Clinical Cancer Research Vol. 5, 95-109, January 1999
© 1999 American Association for Cancer Research

Clinical Trials

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)1

Linda M. Bavisotto, Nilesh H. Patel, Sandra J. Althaus, Douglas M. Coldwell, Hanh V. Nghiem, Tove Thompson, Barry Storer and Charles R. Thomas, Jr.2

Division of Oncology, Departments of Medicine [L. M. B.] and Radiology [N. H. P., S. J. A., D. M. C., H. V. N.], University of Washington Medical Center, Seattle, Washington 98195; Puget Sound Oncology Consortium, Seattle, Washington 98195 [L. M. B., T. T., C. R. T.]; Clinical Statistics, Clinical Division, Fred Hutchinson Cancer Research Center [B. S.], Seattle, Washington 98195; and Department of Radiation Oncology, Medical University of South Carolina/Hollings Cancer Center, Digestive Disease Center, Charleston, South Carolina 29425 [C. R. T.]


   ABSTRACT
Top
 ABSTRACT
INTRODUCTION
 PATIENTS AND METHODS
 RESULTS
 DISCUSSION
 REFERENCES
 
We assessed a regimen of alternating regional and systemic therapy in patients with gastrointestinal malignancies with liver-dominant metastases for feasibility, toxicity, response rate, response duration, patterns of progression, and progression-free and overall survival. Regional therapy comprised selective hepatic transcatheter arterial chemoembolization (TACE) using a suspension of cisplatin and particulate polyvinyl alcohol. This procedure was delivered between cycles of protracted continuous infusion 5-fluorouracil (PCI-5FU) as systemic chemotherapy. Patient eligibility criteria included: (a) having histologically documented adenocarcinoma arising from a gastrointestinal primary site with unresectable liver metastases bidimensionally measurable on computerized tomography scan; (b) age greater than 18 years; and (c) performance status 0–2 (Zubrod). PCI-5FU (250 mg/m2/day) was administered i.v. for 28 days, followed by the first TACE (TACE 1) delivered to the hepatic artery supplying the lobe with the greatest tumor burden. Restaging was performed before TACE 2 and TACE 3, which followed at monthly intervals. PCI-5FU for 21 days was sandwiched between each of the TACE treatments. After the final TACE, maintenance PCI-5FU was given for 28 days of each 35-day cycle until toxicity or progression. Between December 23, 1991, and January 19, 1995, 32 patients were registered in this trial, of whom 27 were eligible; 20 completed one or more treatment cycles and were evaluable for radiographic response. Patients with colorectal liver metastases predominated (74%). Twelve (44%) of 27 patients had failed one or more prior treatment regimens. There were no treatment-related deaths, and hematological and hepatic toxicities were generally manageable and reversible. Two patients, however, developed hepatic abscesses requiring drainage, and one patient developed an infarcted gallbladder, which necessitated cholecystectomy. There were no patients with complete responses; there were 8 (40%) with partial responses, 4 (20%) with minor responses, 2 (10%) with stable disease, and 6 (30%) who progressed on the treatment. The median duration of response for partial responders was 4.2 months (127 days; range, 56–245 days). The median reduction in carcinoembryonic antigen for responders was 87.5%. Two patients underwent subsequent resection of residual metastases; one of them is still alive at 58.4 months follow-up. The predominant site of disease progression was the liver; 25% of the patients progressed in extrahepatic sites. The median overall survival for the whole group is 14.3 months (95% confidence interval, 7.2–16.2). Actuarial overall survival for the whole group at 1 year and 2 years is 57 and 19%, respectively. Alternating systemic PCI-5FU and regional TACE (cisplatin/polyvinyl alcohol) is an active and feasible regimen with manageable toxicities in patients with metastatic gastrointestinal malignancies with liver-dominant disease and merits further investigation. The complications seen were in line with those reported at other specialized centers.


   INTRODUCTION
Top
 ABSTRACT
 INTRODUCTION
PATIENTS AND METHODS
 RESULTS
 DISCUSSION
 REFERENCES
 
Over 200,000 GI3 malignancies are diagnosed annually in the United States, and the liver is the predominant site of metastasis in most of them. Colorectal carcinoma is the third leading cause of cancer-related death in men and women in North America, with an incidence of 160,000 cases and approximately 60,000 deaths annually (1 , 2) . Liver metastases are present at the time of initial diagnosis of colorectal cancer in 20% of patients and are the only site of metastatic disease in one-half of these (3) . Approximately 60% of all patients who die of colorectal cancer will develop liver metastases during the course of their disease (3 , 4) . For patients with colorectal cancer metastases confined to the liver, surgical resection where amenable is the only potentially curative treatment option, but only a small fraction of patients will be candidates for surgery. The survival rate for patients with colorectal cancer after liver metastasectomy with curative intent is 23–65% at 3 years and 25–45% at 5 years (5 , 6) . For patients with GI malignancies whose liver metastases are considered unresectable, there remains a compelling need to develop treatments with higher efficacy.

Little has been published on the results of hepatic-directed therapy in the setting of liver metastases from GI adenocarcinomas other than those of colorectal primary (7) . The metastasectomy experience does not generalize easily to other GI adenocarcinomas (7) , which have distinctive natural histories. Chemoresponsiveness is generally lower, and projected survival for advanced disease generally shorter, than for colorectal cancer. Although promising options for second-line therapy in 5FU-refractory disease have only recently begun to expand for colorectal cancer (8, 9, 10) , there are fewer options for other primary GI adenocarcinomas with liver-dominant metastases. For this reason, a treatment such as chemoembolization is of interest in that its effectiveness may rely more on the dependence of a given tumor on its vascular supply than strictly on its chemosensitivity. Moreover, it may be possible to overcome steep dose-response curves for chemotherapy effect by attaining more protracted and locally concentrated levels of chemotherapeutic agents near the tumor.

Liver metastases have been shown to depend heavily on the hepatic artery for most of their blood supply, whereas the normal liver parenchyma derives the majority of its blood supply from the portal vein (11) . Exploitation of the differential dependence on the hepatic artery between tumor and normal parenchyma has allowed the development of regional treatment strategies, including direct HAI of chemotherapy and various methods of reversible or irreversible vascular occlusion. The historical development and optimization of these regional techniques have been recently reviewed (1 , 7 , 12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 23, 24, 25, 26, 27, 28) .

Hepatic artery occlusion has been studied in depth and its physiological consequences explored (1 , 29, 30, 31) . Ligation of the hepatic artery results in rapid formation of collaterals. Smaller branches can be embolized for more prolonged tumor ischemia. Chemoembolization increases the local concentration of regionally delivered chemotherapeutic agents up to 20-fold and prolongs local dwell time (1) , which may increase the activity of drugs that are marginally active at concentrations achievable with systemic delivery. In the case of drugs with significant first-pass hepatic extraction, hepatic arterial chemoembolization can minimize otherwise limiting systemic toxicities. The embolic particles travel preferentially to the highest flow areas, thereby shunting blood flow to previously less well-perfused areas and increasing drug concentration in hypoxic tumor areas. Anoxia-induced increase in vascular permeability and reperfusion injury may augment local toxicity synergistically (32 , 33) .

Chemoembolization has utility in the treatment of unresectable hepatocellular carcinomas (1 , 20 , 21 , 23 , 34, 35, 36, 37, 38, 39, 40) both for palliation and for neoadjuvant treatment of borderline unresectable tumors with the aim of downstaging for potentially curative resection. HAE has a well-documented role in the palliation of neuroendocrine tumors metastatic to liver (1 , 22 , 41) , which results in improved performance status and reduction of symptoms from vasoactive hormone release. Both of these hypervascular tumors types, as well as other hypervascular malignancies that tend to have isolated liver metastases such as ocular melanoma and GI sarcoma, have occasionally shown marked regression with chemoembolization (1 , 17 , 42) . Less is known about the RRs of hypovascular tumors such as hepatic metastases from adenocarcinomas of various primary sites.

The relationship between RRs and survival in colorectal cancer appears to be complex in that the measures that have significantly enhanced RRs have often failed to translate into clinically significant gains in median survival. Conventional systemic i.v. chemotherapy with single agent 5FU in advanced colorectal cancer has produced RRs of only 7–20% with a median survival of 25–55 weeks (43) . Modulation with folinic acid (leucovorin) has boosted tumor RRs (16–45%) but has produced only marginal survival benefit (2 , 43) . PCI-5FU produced a RR of 22% versus 14% (P = 0.0002) for bolus 5FU regimens in a recent meta-analysis of the Phase III randomized trials that compared these schedules (44) and produced a slight but statistically significant increase in OS. Median survivals were close for both arms, however, at 12.1 months for PCI-5FU versus 11.3 months for bolus 5FU, with similar response durations of 7.1 versus 6.7 months, respectively. To its advantage, PCI-5FU delivers a higher dose-intensity, has produced responses in patients who had progressed on prior 5FU/leucovorin regimens, and is well-tolerated with less hematological toxicity than seen with other delivery schedules (44 , 45) . In addition, a significant prolongation of TTP of disease has also been reported with PCI-5FU compared with bolus 5FU regimens in advanced colorectal cancer (46) .

In terms of regional therapy, HAI of FUdR for liver-dominant metastatic colorectal cancer produces uniformly superior RRs of 40–60%, but in individual trials (47, 48, 49, 50, 51) , it eluded proof of survival benefit. The Meta-Analysis in Cancer Group (52) found that the overall RR for HAI was 41% versus 14% for systemic chemotherapy, with an OR of 0.25 (CI, 0.16–0.40; P < 0.0000000001). Despite this highly significant result, a survival benefit for HAI was achieved only when all of the trials including untreated controls were included in the analysis (P = .0009) but not when trials comparing only HAI and systemic chemotherapy were analyzed (P = 0.14; Ref. 52 ). Another meta-analysis using a different method (53) found that HAI with FUdR produced a modest 10% (P = 0.041) and 6% (P = 0.124) survival advantage at 1 and 2 years, respectively, over systemic chemotherapy.

On the basis of prior experience at our institution with the safety and efficacy of particulate PVA (Ivalon) as a nondegradable embolizing agent (15 , 54) , we elected to use this agent in conjunction with CDDP as the chemoembolizing mixture for this study. Given systemically, CDDP has activity in combination with 5FU in gastric and esophageal carcinomas. A number of clinical trials have explored the utility of adding CDDP in low, weekly doses or in standard doses to i.v. bolus or infusional schedules of 5FU in advanced colorectal cancer, showing a trend toward increased RRs but without prolonging survival (46 , 55) . The concentration of CDDP in the liver can be increased up to 20-fold by hepatic intra-arterial infusion, and the safety, toxicity, and pharmacokinetics of hepatic intra-arterial CDDP administration have been well described (38 , 56, 57, 58) .

Although a minority of patients with advanced colorectal cancer may have isolated liver metastases without detectable extrahepatic disease for extended periods of time, it is not possible to determine a priori which patients will fall into this category. Given the systemic nature of the disease, it stands to reason that a significant proportion of patients treated with regional therapy alone will demonstrate predictable failure at extra-hepatic sites (59) . Accordingly, we elected to alternate regional therapy for liver-dominant disease with ongoing systemic therapy using PCI-5FU to retard the development or progression of failure in extrahepatic sites, hypothesizing that the addition of systemic chemotherapy might favorably impact survival and PFS relative to regional treatment alone.


   PATIENTS AND METHODS
Top
 ABSTRACT
 INTRODUCTION
 PATIENTS AND METHODS
RESULTS
 DISCUSSION
 REFERENCES
 
The design of this trial was approved by the Institutional Review Board of the Fred Hutchinson Cancer Research Center and the Human Subjects Committee of the University of Washington Medical (both in Seattle). All patients gave written, informed consent before treatment.

Patient Selection.
Eligible patients were required: (a) to have histologically verified adenocarcinoma from a GI primary site with unresectable liver metastases measurable on cross-sectional imaging (CT); (b) to have good performance status (SWOG 0–2); (c) to have a life expectancy >8 weeks; and (d) to be at least 18 years old. Patients may have received prior systemic chemotherapy or immunotherapy but could not have received prior 5FU by continuous i.v. infusion, prior hepatic artery or portal vein chemotherapy infusion, or prior HAE or chemoembolization. Small volume (nondominant) extrahepatic metastases were permitted provided the most immediately life-threatening disease was in the liver. Pretreatment laboratory studies required: (a) a WBC >4000/µl; (b) platelets >100,000/µl; (c) serum creatinine <2.0; and (d) one of the following: (a) a prestudy CT scan showing <50% of the liver involved with metastatic disease; or (b) total bilirubin <2.0 mg/dl, aspartate aminotransferase <100 mg/dl, and lactate dehydrogenase <450 mg/dl. Patients with ascites, hepatic encephalopathy, prior hepatic irradiation, poor nutritional risk, active infections, active gastric ulcer, pyschiatric risk, or other illnesses that would preclude the safe administration of chemotherapy or chemoembolization were excluded, as were pregnant or lactating women.

Treatment with Systemic PCI-5FU.
Induction chemotherapy consisting of 5FU 250 mg/m2/day as a protracted continuous i.v. infusion (PCI-5FU) was administered by ambulatory infusion pump via an indwelling central venous access device for 28 days before the first chemoembolization (TACE 1). Pyridoxine (50 mg p.o. t.i.d.) was given while receiving 5FU. Modifications were made in the event of toxicity development as follows:

(a) for Grade 2 toxicity from palmar-plantar erythrodysesthesia (hand-foot syndrome), 5FU was withheld until the toxicity resolved and then resumed at full dose; and
(b) for Grade 3–4 toxicity from hand-foot syndrome, 5FU was withheld until the toxicity resolved and then was restarted with dose reduction to 150 mg/m2/day.
There was no dose reduction for hematological toxicity, but if granulocytes dropped below 1,000/µl or platelets below 50,000/µl, 5FU was withheld until these counts recovered to those levels. Development of angina-like chest pain was a contraindication to further 5FU administration.

PCI-5FU at the same dosage was also administered between TACE procedures, was resumed on hospital discharge, and continued for 21 days until the following TACE. After the final TACE, maintenance chemotherapy with PCI-5FU was administered for 28 days of each 35-day cycle (4 weeks on, 1 week off) until there was evidence of progression.

Regional Selective Hepatic TACE Procedure.
Chemoembolization procedures on all patients were performed by one group of investigators (D. M. C., N. H. P., S. J. A.) in the Interventional Radiology suite at the University of Washington Medical Center. Patients arrived early on the day of the procedure to receive a liter of i.v. prehydration with 5% dextrose in normal saline supplemented with potassium chloride. Antibiotic prophylaxis was not standardized by protocol but was given at the discretion of the medical oncologist or the interventional radiologist performing the procedure. Patients received a celiac plexus block immediately before the procedure, which has been shown to significantly reduce the pain incurred as part of the postembolization syndrome (60 , 61) . The TACE procedure was modeled after the original HAE procedure of Chuang et al. (62) and modified from the procedure previously used at our institution (15 , 54) . Via a transfemoral approach using the Seldinger technique (63) , celiac and superior mesenteric arteriograms were performed using a selective catheter to ascertain any variant arterial visceral anatomy and to document portal vein patency. Coaxially through the selective catheter, a microcatheter was placed into the right or left hepatic artery trunk, where another arteriogram was performed to precisely delineate the distribution of all of the vessels feeding the tumor and to determine optimal placement of the microcatheter tip for embolization. In a few cases where applicable, the internal mammary artery was also selected to determine whether its branches supplied any anterior subcapsular tumor metastases in the liver (64) . When the catheter was within the origin of the artery, 3–6 ml of nonabsorbable PVA foam particles (Ivalon, Contour Emboli, San Francisco, CA)—measuring 150–250 µm in diameter and suspended in dilute nonionic contrast—were injected through the microcatheter in aliquots of approximately 1–3 ml until the slowing of the flow was fluoroscopically evident. A CDDP suspension for intra-arterial injection was used wherein 10 ml of saline was used to suspend particles of a 50-mg vial and injected slowly in 1–3-ml aliquots. (The first three patients on the study received a total administered CDDP dose of 75 mg/m2; in the absence of undue toxicity, the dose for subsequent patients was escalated as planned to 90 mg/m2.) The CDDP infusion was followed with further PVA particle embolization until near-complete stasis of the selected hepatic artery, documented by hand-injection of a contrast. All of the catheters were then removed, and manual digital compression applied over the puncture access site until hemostasis was achieved. Each patient was admitted to the hospital after completion of the procedure, maintained at strict bedrest, and his/her femoral arterial puncture site and distal pulses were monitored by the nursing staff for 6 h postprocedure. After each TACE procedure, patients were monitored for the anticipated postembolization syndrome of fever, pain (abdomen, right upper quadrant, and right shoulder), transient leukocytosis, and elevated transaminases, as well as any other potential complication. Pain was managed with PCA pumps using morphine sulfate or hydromorphone (61) . Patients were discharged when pain could be managed with oral analgesics, when able to take adequate oral fluids, and when serum transaminases were beginning to decline (65) , usually within 2–3 days. Other supportive measures during hospitalization included the regular administration of an H2-blocker (ranitidine or nizatidine) as well as the administration of dexamethasone, ondansetron, and lorazepam as antiemetics before and after CDDP TACE.

Chemoembolization was repeated at monthly intervals for a maximum of two embolizations for unilobar involvement and three embolizations for bilobar involvement. The initial chemoembolization was delivered to the hepatic lobe containing the largest tumor burden. The second chemoembolization was given 1 month later to the same lobe for unilobar and opposite lobe for bilobar involvement; and for patients with bilobar involvement, a third chemoembolization that treated both lobes of the liver was delivered 1 month thereafter. A reduced dose of CDDP (50 mg/m2 was given if at the nadir of the previous cycle, granulocytes dropped <1,000/µl or platelets <70,000/µl, or if the serum creatinine, drawn before the planned TACE, was elevated in the range of 1.6–2.0 mg/dl. If the serum creatinine exceeded 2.0 mg/dl, TACE was delayed until the serum creatinine dropped below this value.

Patient Monitoring.
Serum liver transaminases, total bilirubin, alkaline phosphatase, lactic dehydrogenase, CBC and platelet count, protime, and activated partial thromboplastin time were monitored for each patient just before and 24 h after each TACE procedure. Liver transaminases and WBC were followed daily until hospital discharge. CEA was measured on study entry, before each TACE procedure, and then every 2 months until the patient was off the study.

Restaging with abdominal CT scan, chest radiograph, and physical examination was carried out just before TACE 2 and TACE 3 (weeks 10 and 14). The objective status was recorded at each evaluation. Patients were removed from the study for hepatic or extrahepatic disease progression, for toxicity unacceptable to either the patient or the physician, or if for any reason the patient and/or physician decided it is was in the patient’s best interest to withdraw from the study. All patients were followed until death for PFS and OS.

Definition of Response Criteria and Survival End Points.
Three of the authors (H. V. N., C. R. T., L. M. B.) together reviewed and coded responses for all of the CT scans. Radiographic (CT) responses were judged by standard criteria as follows:

(a) CR indicated complete disappearance of all measurable and evaluable disease, no new lesions, and normalization of tumor markers;
(b) PR indicated a greater than 50% reduction under baseline in the sum of the products of the longest perpendicular diameters of all measurable lesions, no new lesions, and no progression of evaluable disease; and
(c) MR indicated a decrease greater than 25% but less than 50% under baseline in the sum of the products of the perpendicular diameters of all measurable lesions.
Progression was coded for the following circumstances: (a) an increase of 50% or of 10 cm2 (which ever is smaller) in the sum of the products of the perpendicular diameters of the measurable lesions over the smallest sum observed (over baseline if no decrease); (b) the reappearance of any lesions which had disappeared; (c) a clear worsening of any evaluable disease; (d) the appearance of any new lesion; and (e) the failure to return for evaluation due to deteriorating condition (unless clearly due to causes other than the cancer). Stable disease was coded for responses not qualifying for CR, PR, MR, or progression.

Duration of response was defined as the interval from the date that tumors first met the criteria for CR or PR to the date that the disease progression was first detected. PFS and OS were measured from the date of study entry to progression or death.

Statistical Analysis.
The primary study end point was the objective radiographic (CT) RR; secondary end points were PFS status and OS. RRs were determined in two ways, with both results given. In the first determination, all of the eligible patients without exclusion, including those who received no or less than one complete cycle of the planned treatment or had protocol violations, were included in the denominator for RR according to intent-to-treat principles; patients not assessable for radiographic response or inevaluable due to protocol violation were counted as treatment failures. A parallel analysis including only protocol-adherent patients who completed one or more cycles of planned treatment and were, therefore, evaluable for radiographic response is given for comparison.

Pre- and posttreatment CEA tumor marker levels were recorded as another indicator of tumor response for each subject. Mean values of pre- and posttreatment CEA levels were compared using a two-sided Student’s t test for paired data for both radiographic responders and nonresponders.

PFS and OS were estimated using the product-limits method of Kaplan and Meier (66) .


   RESULTS
Top
 ABSTRACT
 INTRODUCTION
 PATIENTS AND METHODS
 RESULTS
DISCUSSION
 REFERENCES
 
Patient and Tumor Characteristics.
Between December 23, 1991, and January 19, 1995, the Puget Sound Oncology Consortium registered 32 patients from 13 participating institutions into this clinical trial coordinated through the University of Washington Medical Center, with follow-up through December 1997. Patient characteristics and primary disease sites are summarized in Tables 1Citation and 2Citation . Median age was 57 years (range, 38–77), and males outnumbered females 2:1. The majority—20 (74%) of 27—had a primary adenocarcinoma of the colon or rectum; 2 patients each had gastric and pancreatic carcinoma, 1 patient each had esophageal, intrahepatic bile duct, and gallbladder primary sites. Twelve (44%) of 27 had progressed on one or more prior treatment regimens. Six had unilobar hepatic involvement with metastases; and in 21, the involvement was bilobar. Eleven (40.7%) of 27 patients had small-volume extrahepatic metastases on entering the study. All but one patient had performance status 0–1. The number of chemoembolizations completed is also shown in Table 1Citation , with 17 patients completing two or more TACEs.


View this table:
[in this window]
[in a new window]
  		Table 1 Patient characteristics

 

View this table:
[in this window]
[in a new window]
  		Table 2 Tumor characteristics

 
Of 27 eligible patients in our trial, 7 were inevaluable for response, although all of them were included in an analysis of RR using intent-to-treat principles. Two of the seven patients lacked appropriate CT documentation to assess response, and five received less than one complete cycle of therapy. This rate of early withdrawal from our study was higher than we would have liked but on further examination did not appear to be due to excessive toxicity from the initial 28-day PCI-5FU systemic chemotherapy nor to rapid/early progression of disease on study entry. Rather, several patients declined further participation on receiving decisions of noncoverage from their insurers for TACE procedures and the associated hospital admissions; one patient declined, stating concerns over anticipated toxicity from the TACE procedure after starting on PCI-5FU, and one patient who was enrolled and started on PCI-5FU at a remote location subsequently refused the first TACE procedure.

Toxicity.
Twenty-six patients with a total of 48 TACE procedures were evaluable for toxicity. There were no treatment-related deaths. Of the 26 patients evaluable for toxicity, 21 (81%) experienced at least one grade 3 event, and 8 (31%) experienced a grade 4 event. The incidence of grade 3 and grade 4 toxic events is summarized in Table 3Citation . Palmar-plantar erythrodysesthesia (hand-foot syndrome) was responsible for delays and dosage reductions in a minority of patients while receiving PCI-5FU (one grade 3 and three grade 4 events). One patient developed grade 4 diarrhea. After TACE procedures, nearly all of the patients experienced signs and symptoms compatible with postembolization syndrome, namely fever, pain, transient leukocytosis, and elevated transaminases. This was especially pronounced after the first TACE procedure and often less prominent after subsequent chemoembolizations. Fatigue was a frequent accompaniment, particularly after the initial TACE procedure. Nausea and vomiting, despite premedication and ongoing prophylaxis with ondansetron and dexamethasone, were common. Hematological toxicities were infrequent. Two patients developed grade 4 thrombocytopenia (lowest platelet count = 22,000/µl 1 week after TACE 1), and two patients had grade 3 granulocytopenia without related complication. Nephrotoxicity or peripheral neurotoxicity potentially related to systemic CDDP exposure were not seen; only one patient had a highest serum creatinine greater than 1.4 mg/dl.


View this table:
[in this window]
[in a new window]
  		Table 3 Incidence of grade 3–4 toxic events (n = 26 patients; 48 TACE procedures evaluable)

 
Pain ratings were lower in the first 24 h post-TACE in patients who had successful celiac plexus blocks, but in all of the cases, pain was well managed with opioid PCA pumps. Pain ratings were frequently lower after subsequent TACE procedures than after the initial TACE.

Ten patients (38%) experienced one or more of the treatment-related complications listed in Table 4Citation . One patient (4%) with colon cancer developed an infarcted gallbladder after TACE 2 and required a cholecystectomy. The patient recovered fully. Two patients (8%) developed hepatic abscesses. One of these, a patient with colon cancer, developed a liver abscess that was culture-negative after percutaneous drainage but grew Klebsiella pneumoniae in blood cultures. This patient recovered fully with antibiotic treatment. The second patient, who had pancreatic cancer metastatic to the liver, developed a liver abscess that grew enterococcus and Clostridium perfringens in cultures obtained by percutaneous drainage, and blood cultures grew K. pneumoniae. This patient suffered grade 4 hypotension and hyperglycemia in the setting of sepsis and mild baseline pancreatic insufficiency but recovered fully with percutaneous drainage and antibiotic treatment. Two patients developed deep venous thrombosis related to central venous catheters and were treated with anticoagulation. One patient with colon cancer received chest tube drainage for a unilateral pleural effusion after TACE 2, symptomatic with dyspnea at rest. These symptoms did not recur.


View this table:
[in this window]
[in a new window]
  		Table 4 Complications (n = 26 patients)

 
RR.
Best responses on study and response durations are shown in Table 5Citation .


View this table:
[in this window]
[in a new window]
  		Table 5 Best response on-study and response durations

 
When only those patients who completed one or more of the planned treatment cycles and were, therefore, evaluable for radiographic response were included in the determination of RR (n = 20, see Table 5ACitation ), there were no complete responders, 8 (40%) partial responders, 4 (20%) patients with MRs, 2 (10%) with stable disease, and 6 (30%) who progressed on treatment. In a parallel determination of RR according to intent-to-treat principles, all eligible patients without exclusion were included in the denominator for RR, and those with protocol violations or those who received less than one cycle of planned treatment and were, therefore, technically unassessable for radiographic response were counted as treatment failures (n = 27; see Table 5BCitation ). In this analysis, the PR and MR rates are slightly decreased to 30 and 15%, respectively. The median response duration for the partial responders was 4.2 months (range, 56–245 days). Representative CT scans from two of the partial responders are shown in Fig. 1Citation .



View larger version (111K):
[in this window]
[in a new window]
  		Fig. 1. Representative CT scans from two partial responders before treatment (left panels) and after treatment (right panels). The patient whose scans are shown in the upper panels had 95% reduction in CEA; the patient whose scans are shown in the lower panels had 75% reduction in CEA.

 
The median on-study reduction in CEA levels in radiographic responders was 87.5%. Changes in CEA with each treatment cycle are shown graphically in Fig. 2Citation for radiographic responders and nonresponders. Pre- and posttreatment mean CEA values were significantly different (P < 0.05 by two-tailed Student’s t test for paired data) for radiographic responders but not for nonresponders.



View larger version (18K):
[in this window]
[in a new window]
  		Fig. 2. Log-linear scatter plots of change in CEA levels in units/ml over time after the baseline (prestudy) assessment. The X-axis indicates time intervals between TACE treatments, generally one month apart. Upper panel, CEA values for all of the patients with these data available at the time of each assessment. The mean CEA value for each assessment time is also shown. —-, connects mean values for radiographic responders; - - -, connects mean values for radiographic nonresponder. The median reduction in CEA for radiographic responders was 87.5%. Comparing the mean prestudy CEA value with the mean CEA value obtained just before the final TACE using a two-tailed Student’s t test for paired data, a significant difference (P < .05) was shown for responders but not for nonresponders. Lower panel, CEA values at the time of each assessment are compared longitudinally for individual patients for both radiographic responders (left, —-) and nonresponders (right, - - -).

 
In cases where CEA was informative as a tumor marker, the decline in CEA generally paralleled or slightly preceded the radiographic reduction in lesion size for a given patient. There were interesting exceptions, however, as in the case of one patient who had a significant drop with near-normalization of CEA—a 93% reduction from 95.2 to 5.8 units/ml over the course of three TACE procedures–but who did not achieve sufficient radiographic reduction in lesion size to qualify for a MR. Development of a tiny new abnormal but indeterminate area in the liver led to the patient’s being taken off the study and coded as progression (in fact, this may have been a small parenchymal infarct, although new metastasis could not be excluded; see "Discussion.") Although the prestudy evaluation for this patient had deemed her liver metastases unresectable, she went on to further radiolabeled monoclonal antibody treatment and subsequently underwent trisegmentectomy for resection of residual metastases. She is alive and without evidence of recurrence at 58.4 months of follow-up.

A second patient also was recommended for metastasectomy of residual hepatic lesions after maximal response on study. This patient, whose metastatic lesions were confined to the left hepatic lobe, was considered unresectable prestudy because of the proximity of metastases to the middle hepatic vein. She experienced modest radiographic tumor shrinkage on-study, although insufficient to qualify by the criteria for a MR. On clinical reassessment after TACE 3 with CT arteriogram and chest radiograph, and with intraoperative ultrasound failing to show any new liver lesions, the possibility of achieving a reasonable surgical margin was suggested, and she went on to have a left hepatic lobectomy. Unfortunately, the surgical resection specimen revealed microscopic tumor abutting the middle hepatic vein; therefore, a clear margin was not achieved. This patient died at 9.1 months of follow-up.

Sites of Progression.
Table 6Citation summarizes the first site of disease progression—categorized as liver only, extrahepatic only, or both—that was responsible for each patient being taken off-study. The table is further categorized according to the patient’s metastatic status on study entry (liver only, or liver and extrahepatic). The upper panel shows the sites of progression for all of the radiographically evaluable patients (n = 20), and the bottom panel shows the sites of progression for patients only with radiographically-responsive disease in the study (responders with either PR or MR, n = 12).


View this table:
[in this window]
[in a new window]
  		Table 6 Sites of progression

 
Of the 16 patients without detectable extrahepatic metastases prestudy, 3 were not evaluable radiographically for progression. Of the 11 patients who had small-volume extrahepatic disease prestudy, 4 were not evaluable for progression (1 progressed and went off the study before receiving a cycle of therapy, 3 refused therapy and/or had protocol violations). For all 20 radiographically evaluable patients, the predominant site of disease progression was in the liver. One-fourth of the patients (5 of 20) showed progression in an extrahepatic site: 3 progressed exclusively outside of the liver, and 2 progressed simultaneously in the liver and an extrahepatic site. Regarding these five patients who demonstrated progression in an extrahepatic site, three had no detectable prestudy extrahepatic disease. The remaining two patients who progressed in extrahepatic sites had pre-existing minimal extrahepatic disease on study entry.

TTP, PFS, and OS.
All of the 27 eligible patients were evaluable for survival. All but one of these patients have died. The median OS for the whole group is 14.3 months (95% confidence interval, 7.2–16.2), whereas the median PFS is 4.4 months (95% confidence interval, 3.3–6.8; see Fig. 3Citation . Actuarial OS is 57% at 1 year, and 19% at 2 years.



View larger version (15K):
[in this window]
[in a new window]
  		Fig. 3. Upper panel, Kaplan-Meier curve showing cumulative OS for all of the 27 eligible patients (median survival, 14.3 months). Lower panel, TTP for the same group (median TTP, 4.4 months).

 

   DISCUSSION
Top
 ABSTRACT
 INTRODUCTION
 PATIENTS AND METHODS
 RESULTS
 DISCUSSION
REFERENCES
 
We compare our results to those of other hepatic TACE regimens described in the literature for patients with unresectable hepatic metastases from colorectal cancer (13 , 17 , 21 , 33 , 67, 68, 69, 70, 71, 72, 73, 74, 75, 76, 77, 78, 79) , as summarized in Table 7Citation , none of which added concurrent, sequential, or alternating systemic chemotherapy to chemoembolization as our trial did. Although it is difficult to make meaningful interpretations because of the heterogeneous patient selection criteria, small trial sizes, and varying techniques and agents for chemoembolization, some general impressions can be drawn from a review of these studies. Many trials (33 , 69 , 72, 73, 74) have required that enrolled patients have progressed on prior front-line chemotherapy; exclusion of patients with extrahepatic metastases was variable. Most studies showed a median survival of 10–13 months. CRs were occasionally seen (68 , 71) . A study from the European Organization for Research and Treatment of Cancer observed responses in hypovascular as well as hypervascular tumors (68) and reported an improved performance status in patients treated with chemoembolization. Early attempts at whole liver embolization resulted in hepatic insufficiency and extreme, often protracted, fatigue (69 , 80) leading to a recommendation for staged, sequential unilobar embolizations at intervals of 4–6 weeks (80) and limiting the volume of the embolizing mixture. To determine which patients might derive the most benefit from chemoembolization, multivariate regression analysis identified several prognostic factors associated with improved survival, including a good performance status and the absence of extrahepatic metastases (74) and uniform radiographic distribution of the embolizing mixture throughout the metastatic lesions (71) .


View this table:
[in this window]
[in a new window]
  		Table 7 Results of selected reported chemoembolization trials in colorectal cancer metastatic to liver

 
The complication of gallbladder infarction and, to a lesser degree, the symptoms suggestive of gastritis seen in our study are presumably related to inadvertent misperfusion or reflux of embolic materials into the cystic artery or gastroduodenal artery, respectively. Gallbladder infarction has been a noted complication in other HAE or chemoembolization trials (20 , 21 , 81, 82, 83) . Improvements in technique have reduced the morbidity associated with chemoembolization since the earliest reported trials, as increasing experience with the chemoembolization procedure at specialized centers has decreased the rate of inadvertent misperfusions (15 , 82 , 83) . Hepatic abscess with sepsis, seen in two patients in our study, is also a previously described risk of chemoembolization (81 , 84 , 85) . In hind sight, optimal antibiotic prophylaxis was not used in our trial, and it is likely that the septic complications in our trial would have been reduced or eliminated if antibiotic prophylaxis had been standardized by protocol. The adoption of extended and often broad-spectrum parenteral antibiotic prophylaxis (13 , 20 , 21 , 27 , 72 , 80 , 81 , 84 , 85) , routine blood culturing, and in some cases a bowel prep before the chemoembolization procedure (33) have decreased the complication rates of hepatic abscesses and sepsis reported by a number of centers, particularly where microfibrillar bovine collagen (Angiostat) was used as the embolizing agent.

When compared with the historical data from TACE trials in colorectal cancer, none of which added systemic chemotherapy, our median survival of 14.3 months does compare favorably (is in the higher end of the published range of median survivals given in Table 7Citation ), despite some patient selection features in our trial that could be expected to produce adverse survival effects. Although the majority of patients in our trial (74%) had colorectal cancer, the inclusion of patients with other primary GI adenocarcinomas metastatic to liver may dilute the RR and lower the median OS somewhat when compared with trials looking at colorectal cancer alone. In addition, several patients in our study had unresectable primary tumors (in each case, liver metastases dominated the clinical picture), which might be expected to adversely affect survival. It is not completely clear whether the addition of systemic chemotherapy in our trial may have helped to offset the potential adverse survival effects from these patient selection factors in our trial. Other factors affecting the mix of patients among different trials (the proportion of patients who had progressed on prior therapy, bulky versus early metastatic disease, and the presence or absence of extrahepatic disease on study entry) can certainly be expected to influence the variation in the range of median survivals reported.

Despite reasonable RRs, reported response durations with chemoembolization remain short. Some reasons for this may be the development of drug resistance, formation of collateral vessels, or acceleration of extrahepatic metastatic growth in the absence of concomitant systemic chemotherapy. Hunt et al. (70) observed in a group of patients carefully selected to exclude extrahepatic metastases at presentation that 36% developed extrahepatic disease sites while on the study, which did not include systemic chemotherapy. In an early study by the University of Michigan group, 73% of patients whose metastatic disease was clinically confined to the liver on study entry and who received only regional treatment via hepatic artery infusion of FUdR died of uncontrolled extrahepatic malignant disease (86) . More recently, the University of Michigan group (87) reported results of a trial combining three exclusively regional modalities for 47 patients with diffuse hepatic involvement with either primary hepatobiliary tumors, or metastases from other cancers, mostly colorectal. Concurrent HAI (FUdR plus leucovorin) was administered for radiosensitization along with whole liver radiotherapy, followed by hepatic arterial chemoembolization using mitomycin C and PVA. Hepatocellular carcinomas, neuroendocrine tumors, and metastases of unknown primary tumors showed higher RRs than metastatic colorectal cancers or cholangiocarcinoma. During this trial, 54% of the colorectal cancer patients progressed in an extrahepatic site after delivery of this multimodal regional regimen, which included no systemic chemotherapy.

During our study of alternating systemic and regional treatment, in contrast to the foregoing, only 5 (25%) of 20 patients showed progression in extrahepatic sites, and 2 of these 5 patients had prestudy baseline extrahepatic metastases. This may suggest a benefit from the addition of systemic chemotherapy to our TACE regional regimen in terms of reduction in the proportion of patients failing in extrahepatic sites when compared with the foregoing studies, but given the patient selection (and other) differences between these separate trials, confirmation with larger numbers of patients in a prospective comparative trial would be required to ascertain a benefit from the addition of systemic chemotherapy. In our study, there does not appear to be a major improvement in TTP or OS compared with published hepatic chemoembolization studies in colorectal cancer that did not employ systemic chemotherapy.

Whether the addition of systemic treatment to chemoembolization reduces the incidence of extrahepatic metastases, and indeed the converse, whether the "addition" of TACE to systemic therapy reduces the intrahepatic progression of disease, are compelling questions that, given the limitations of any Phase II trial, our study, as designed, cannot completely answer. In this regard, it will be of interest to learn the results of a pilot Phase II trial recently conducted by the SWOG (SWOG 9051) and coordinated through the University of Southern California-Norris. This trial administered multi-agent whole liver chemoembolization and sequential systemic chemotherapy to patients with advanced colorectal cancer. The study has completed accrual, with results yet to be reported. Recently, Wanebo et al. (59) reported a superior median survival of 18 months using systemic chemotherapy (continuous infusion 5FU plus leucovorin) in alternation with regional therapy (HAI with FUdR plus dexamethasone) with low toxicity in patients with colorectal cancer and dominant liver metastases.

The evaluation of radiographic response by standard criteria may not be adequate for hepatic chemoembolization studies (23 , 33) . The strict criterion defining partial (major) response in our study (>=50% reduction in lesion size) may have excluded some patients with clinically meaningful responses. The previously noted patient in our study who is still alive at 58.4 months follow-up after poststudy resection of residual liver metastases by trisegmentectomy illustrates this consideration. Although it cannot be definitively claimed that a resectable status was achieved as a result of therapy from this study (as she went on to have additional treatment before resection), her significant on-study decrease in CEA, despite the relatively minor associated radiographic changes, suggests a biological response that may have contributed significantly to her ability to undergo subsequent resection. Venook et al. (23) found liquefaction necrosis both radiographically and pathologically after hepatic chemoembolization for hepatocellular carcinoma, and observed that radiographic response may underestimate true response, because of the constraints on tumor shrinkage of a background of edema and necrosis.

In this regard, at least two groups (33 , 79) have used a different set of criteria to define radiographic response in their respective hepatic chemoembolization studies. Responses were defined by either a decrease in density on CT scan in at least 75% of a metastatic lesion or a cystic pattern, (both consistent with necrosis) or a >=25% (instead of >=50%) decrease in the size of the lesion without development of concomitant lesions. We did not specifically compare CT attenuation of metastatic lesions in our study, but if we were to apply the more liberal size reduction criterion for response to our study, both the partial and minor responders (12 of 20; 60%) would qualify for major response. Our experience suggests that radiographic response should not be viewed in isolation from other markers of tumor response. To document true response would require biopsy or, less invasively, metabolic imaging by positron emission tomography, as in the small series by Vitola et al. (30) in patients with liver metastases treated with chemoembolization.

Meakem et al. (88) have described the evolution over a period of time in CT appearance of metastatic liver lesions in seven patients after chemoembolization with Angiostat (cross-linked bovine collagen), CDDP, mitomycin C, and doxorubicin, using volumetric rather than cross-sectional area measures. At 1 month after chemoembolization, despite declining CEA levels that suggested response, three patients developed low attenuation regions in areas in which there had been no previous lesion, but in the distribution of the arterial tree. Although the significance of these areas was uncertain, the decreasing CEA levels and subsequent evolution in appearance of these sites on CT with sharper demargination and diminishing size suggested that they were regions of hepatic ischemia/infarction as opposed to heretofore unidentifiable metastases now "unmasked." Follow-up CT scans at 2–3 months after a single chemoembolization revealed maximal effect on tumor volume in their study.

Several unsettled questions include whether multi-drug chemoembolization is superior to single-drug chemoembolization in activity and in delaying the development of drug resistance, and which drug(s) is most effective when administered in conjunction with a hepatic arterial embolizing agent. A promising agent to consider in this regard is the new third-generation platinum analogue oxaliplatin, a member of the 1,2-diaminocyclohexane family of platinum compounds. Whereas conventional systemic chemotherapy trials combining CDDP with various schedules of 5FU proved generally disappointing (46) , oxaliplatin has been recently found to possess distinct first- and second-line activity as a single agent in treating advanced colorectal cancer (9 , 10) and, when given in combination with 5FU and leucovorin, has been shown to increase the objective RR more than 2-fold, while increasing TTP with a gain of 3 months (89) . Hence, it may be advantageous to use oxaliplatin rather than CDDP, alone or in combination, in future chemoembolization studies along with the embolizing agent. Another widely used agent, mitomycin C, has alternative advantages of an enterohepatic recirculation and conversion to more active metabolites in a hypoxic milieu (1 , 20 , 21 , 90 , 91) .

The RR in our trial, even employing the intent-to-treat analysis in which patients who receive less than one cycle of treatment are counted as treatment failures instead of inevaluable for response (30% PR, 15% MR; see Table 5BCitation ), would ordinarily be indicative of an active regimen meriting further exploration, particularly in previously treated patients, in the absence of undue toxicity. Several factors, however, make the straightforward introduction of this regimen into Phase III testing somewhat problematic. Drawbacks of this approach are the technical difficulty, expense, and need for hospitalization after chemoembolization. Given the higher-than expected early withdrawal rate, the risk of procedure-related complications, and the interim development of promising new systemic treatment options, it may be that our study, as designed and analyzed, does not merit further testing currently in the Phase III setting. Prospective randomized clinical trials to further study hepatic chemoembolization may be challenging to perform given the intrinsic clinical heterogeneity of patients with hepatic tumors (7) and the diversity of strategies in the reported Phase II studies. Future trials must continue to be carried out in centers with expertise in the required technical skills to minimize procedure-associated morbidities. Patient accrual may be slower because of the availability of several promising new therapeutic options for 5FU-refractory advanced colorectal cancer since the closure of this trial. Active second-line agents in current clinical trials such as irinotecan, oxaliplatin, and several thymidylate synthase inhibitors will likely compete for patients who, when therapeutic options were more limited, would have been reasonable candidates for chemoembolization. More importantly, these new agents, the availability of which could not have been anticipated at the outset of our trial, are generally less technically complex to administer, do not require hospitalization or PCA for procedure-related pain, and may, therefore, represent less toxic, less expensive, and more convenient treatment alternatives. It would be both ethical and prudent to await early results and toxicity profiles from the major trials underway with these agents, as new treatment algorithms for their usage are evolving, before deciding whether and how to appropriately further test TACE with or without systemic chemotherapy; perhaps this would be in a subset of patients failing to respond to these systemic therapies, where the risk of potential treatment-related complications from the TACE procedure might be justified. In addition to RR and survival measures to assess efficacy, such trials should include quality of life and cost as additional outcome measures.


   ACKNOWLEDGMENTS
 
We thank Fran Chard and Deborah Bassuk at Puget Sound Oncology Consortium for expert assistance with data management and Karen Capps for preparation of the manuscript. We acknowledge Robert Livingston and Patrick Freeny and thank the participating member-investigators of Puget Sound Oncology Consortium and other regional physicians who enrolled patients from the following institutions into this clinical trial: Bishop Clarkston Hospital (Omaha, NE); Everett Clinic (Everett, WA); Evergreen Cancer Center (Kirkland, WA); Internal Medicine Associates (Anchorage, AK); Madrona Medical (Bellingham, WA); Memorial Clinic (Olympia, WA); Oncology Hematology Specialists (Lewiston, ID); Overlake Internal Medicine (Bellevue, WA); St. Mary Regional Cancer Center (Walla Walla, WA); Tacoma General Hospital (Tacoma, WA); University of Washington Medical Center (Seattle, WA); Valley General Medical Center (Auburn, WA); and Western Montana Clinic (Missoula, MT).


   FOOTNOTES
 
The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

1 Early data from this clinical trial were presented in abstract form by C. R. Thomas, Jr., at the American Radium Society meeting on May 3, 1995, Paris, France. Back

2 To whom requests for reprints should be addressed, at Department of Radiation Oncology/Hollings Cancer Center, Digestive Disease Center, Medical University of South Carolina, 171 Ashley Avenue, Charleston, SC 29425. Phone (803) 792-3273; Fax: (803) 792-5498; E-mail: thomas{at}radonc.musc.edu Back

3 The abbreviations used are: GI, gastrointestinal; CDDP, cis-diammine-dichloro-platinum (cisplatin); CEA, carcinoembryonic antigen; CR, complete response; CT, computerized tomography; 5FU, 5-fluorouracil; FUdR, fluorodeoxyuridine (floxuridine); HAE, hepatic artery embolization; OS, overall survival; PCI-5FU, protracted continuous infusion 5FU; PFS, progression-free survival; PR, partial response; PVA, polyvinyl alcohol; RR, response rate; SWOG, Southwest Oncology Group; TACE, transcatheter arterial chemoembolization; TTP, time to progression; HAI, hepatic arterial infusion; MR, minor response; PCA, patient-controlled analgesia. Back

Received 7/ 9/98; revised 9/29/98; accepted 10/ 2/98.


   REFERENCES
Top
 ABSTRACT
 INTRODUCTION
 PATIENTS AND METHODS
 RESULTS
 DISCUSSION
 REFERENCES
 

  1. Alexander H. R., Bartlett D. L., Fraker D. L., Libutti S. K. Regional treatment strategies for unresectable primary or metastatic cancer confined to the liver. Princ. Pract. Oncol. Updates, 10: 1-19, 1996.
  2. Advanced Colorectal Cancer Meta-Analysis Project. Modulation of fluorouracil by leucovorin in patients with advanced colorectal cancer: evidence in terms of response rate. J. Clin. Oncol., 10: 896-903, 1992.[Abstract]
  3. Wagner J. S., Adson M. A., VanHeerden J. A., Adson M. H., Ilstrup D. M. The natural history of hepatic metastases from colorectal cancer. Ann. Surg., 1: 201-209, 1989.
  4. Weiss L., Grundmann E., Torhorst J., Hartveit F., Moberg I., Eder M., Fenoglio-Preiser C. M., Napier J., Horne C. H., Lopez M. J. Haematogenous metastatic patterns in colonic carcinoma: an analysis of 1541 necropsies. J. Pathol., 150: 195-203, 1986.[Medline]
  5. Bismuth H., Adam R., Levi F., Farabos C., Waechter F., Castaing D., Majno P., Engerran L. Resection of nonresectable liver metastases from colorectal cancer after neoadjuvant chemotherapy. Ann. Surg., 224: 509-522, 1996.[Medline]
  6. Nordlinger B., Vaillant J. C., Guiguet M., Balladur P., Paris F., Bachellier P., Jaeck D. Survival benefit of repeat liver resections for recurrent colorectal metastases: 143 cases. Association Francaise de Chirurgie. J. Clin. Oncol., 12: 1491-1496, 1994.[Abstract]
  7. McMurrick P. J., Nelson H. Liver-directed therapies for gastrointestinal malignancies. Curr. Opin. Oncol., 9: 367-372, 1997.[Medline]
  8. Pitot H. C., Wender D. B., O’Connell M. J., Schroeder G., Goldberg R. M., Rubin J., Mailliard J. A., Knost J. A., Ghosh C., Kirschling R. J., Levitt R., Windschitl H. E. Phase II trial of irinotecan in patients with metastatic colorectal carcinoma. J. Clin. Oncol., 15: 2910-2919, 1997.[Abstract]
  9. Ducreux M., Louvet C., Bekradda M., Cvitkovic E. Oxaliplatin for the treatment of advanced colorectal cancer: future directions. Semin. Oncol., 25 (Suppl. 5): 47-53, 1998.
  10. Cvitkovic E. A historical perspective on oxaliplatin: rethinking the role of platinum compounds and learning from near misses. Semin. Oncol., 25 (Suppl 5): 1-3, 1998.
  11. Breedis C., Young G. Blood supply of neoplasms of the liver. Am. J. Pathol., 30: 969-972, 1954.
  12. Campbell K. A., Burns R. C., Sitzmann J. V., Lipsett P. A., Grochow L. B., Niederhuber J. E. Regional chemotherapy devices: effect of experience and anatomy on complications. J. Clin. Oncol., 11: 822-826, 1993.[Abstract/Free Full Text]
  13. Cascinu S., Wadler S. Chemo-embolization in the treatment of liver metastases from colorectal cancer. Cancer Treat. Rev., 22: 355-363, 1996.[Medline]
  14. Choi J. Regional transcatheter therapy of hepatic neoplasms. Cancer Control, 3: 407-420, 1996.[Medline]
  15. Coldwell D. M., Mortimer J. E. Transcatheter therapy for malignant neoplasms. West. J. Med., 151: 299-303, 1989.[Medline]
  16. Coldwell D. M., Stokes K. R., Yakes W. F. Embolotherapy: agents, clinical applications, and techniques. Radiographics, 14: 623-627, 1994.[Abstract]
  17. Lyster M. T., Benson A. B., III, Vogelzang R., Talamonti M. Chemoembolization: alternative for hepatic tumors. Contemporary Oncology, : 17-26, 1993.
  18. Pentecost M. J. Transcatheter treatment of hepatic metastases. Am. J. Radiol., 160: 1171-1175, 1993.[Abstract/Free Full Text]
  19. Rossi P., Bezzi M. Interventional radiology in gastrointestinal neoplasms. Curr. Opin. Oncol., 7: 367-372, 1995.[Medline]
  20. Soulen M. C. Chemoembolization of hepatic malignancies. Oncology (Basel), 8: 77-89, 1994.
  21. Soulen M. C. Chemoembolization of hepatic malignancies. Semin. Interventional Radiol., 14: 305-311, 1997.
  22. Uchida H., Matsuo N., Sakaguchi H., Nagano N., Nishimine K., Ohishi H. Segmental embolotherapy for hepatic cancer: keys to success. Cardiovasc. Interventional Radiol., 16: 67-71, 1993.[Medline]
  23. Venook A. P., Warren R. S. Regional chemotherapy approaches for primary and metastatic liver tumors. Surg. Oncol. Clin. N. Am., 5: 411-427, 1996.[Medline]
  24. Daniels J., Daniels A., Morris M., Groshen S., Lewis B., York M., Barth N. Hepatic artery embolization with Angiostat collagen alone or with cisplatin, Mitomycin C, doxorubicin. Phase I/II evaluation in 525 patients demonstrates survival advantage with addition of chemotherapy. Proc. Am. Soc. Clin. Oncol., 10: 161 1991.
  25. Davidson B. S., Izzo F., Chase J. L., DuBrow R. A., Patt Y., Hohn D. C., Curley S. A. Alternating floxuridine and 5-fluorouracil hepatic arterial chemotherapy for colorectal liver metastases minimizes biliary toxicity. Am. J. Surg., 172: 244-247, 1996.[Medline]
  26. Kemeny N., Seiter K., Conti J. A., Cohen A., Bertino J. R., Sigurdson E. R., Botet J., Chapman D., Mazumdar M., Budd A. J. Hepatic arterial floxuridine and leucovorin for unresectable liver metastases from colorectal carcinoma: new dose schedules and survival update. Cancer (Phila.), 73: 1134-1142, 1994.[Medline]
  27. Patt Y. Z. Regional hepatic arterial chemotherapy for colorectal cancer metastatic to the liver: the controversy continues. J. Clin. Oncol., 11: 815-819, 1993.[Free Full Text]
  28. Stagg R. J., Venook A. P., Chase J. L., Lewis B. J., Warren B. S., Roh M., Mulvihill S. J., Grobman B. J., Rayner A. A., Hohn D. C. Alternating hepatic arterial floxuridine and fluorouracil: a less toxic regimen for the treatment of liver metastases from colorectal cancer. J. Natl. Cancer Inst., 83: 423-426, 1991.[Abstract/Free Full Text]
  29. Taourel P., Dauzat M., Lafortune M., Pradel J., Rossi M., Bruel J-M. Hemodynamic changes after transcatheter arterial embolization of hepatocellular carcinomas. Radiology, 191: 189-192, 1994.[Abstract/Free Full Text]
  30. Vitola J. V., Delbeke D., Meranze S. G., Mazer M. J., Pinson C. W. Positron emission tomography with F-18-fluorodeoxyglucose to evaluate the results of hepatic chemoembolization. Cancer (Phila.), 78: 2216-2222, 1996.[Medline]
  31. Yamashita Y., Takahashi M., Koga Y., Saito R., Nanakawa S., Hatanaka Y., Sato N., Nakashima K., Urata J., Yoshizumi K. Prognostic factors in liver metastases after transcatheter arterial embolization or arterial infusion. Acta Radiol., 31: 269-74, 1990.[Medline]
  32. Wallace S., Carrasco C. H., Charnsangavej C., Richli W. R., Wright K., Gianturco C. Hepatic artery infusion and chemoembolization in the management of liver metastases. Cardiovasc. Interventional Radiol., 13: 153-160, 1990.[Medline]
  33. Tellez C., Benson A. B., III, Lyster M. T., Talamonti M., Shaw J., Braun M. A., Nemcek A. A., Jr., Vogelzang R. L. Phase II trial of chemoembolization for the treatment of metastatic colorectal carcinoma to the liver and review of the literature. Cancer (Phila.), 82: 1250-1259, 1998.[Medline]
  34. Hasuike Y., Okamura J., Furukawa J., Naoi M., Takata N., Maruyama H., Kinuta M., Yayoi E., Oi H., Okamoto S., Monden M., Mori T., Sakurai M. Efficacy of combination treatment—(TAE with Adriamycin and ethanol)—for hepatocellular carcinoma. Cancer Chemother. Pharmacol., 31 (Suppl I): S30-S34, 1992.
  35. Kawai S., Okamura J., Ogawa M., Ohashi Y., Tani M., Inoue J., Kawarada Y., Kusano M., Kubo Y., Kuroda C., Sakata Y., Shimamura Y., Jinno K., Takahashi A., Takayasu K., Tamura K., Nagasue N., Nakanashi Y., Makino M., Masuzawa M., Mikuriya S., Monden M., Yumoto Y., Mori T., Oda T., The Cooperative Study Group for Liver Cancer Treatment of Japan. Prospective and randomized clinical trial for the treatment of hepatocellular carcinoma: a comparison of lipiodol-transcatheter arterial embolization with and without Adriamycin (first cooperative study). Cancer Chemother. Pharmacol., 31 (Suppl. I): S1-S6, 1992.
  36. Trinchet J. C., Rached A. A., Beaugrand M. A comparison of Lipiodol chemoembolization and conservative treatment for unresectable hepatocellular carcinoma. N. Engl. J. Med., 332: 1256-1261, 1995.[Abstract/Free Full Text]
  37. Vetter D., Wenger J-J., Bergier J-M., Doffoel M., Bockel R. Transcatheter oily chemoembolization in the management of advanced hepatocellular carcinoma in cirrhosis: results of a Western comparative study in 60 patients. Hepatology, 13: 427-433, 1991.[Medline]
  38. Yodono H., Saito Y., Saikawa Y., Midorikawa H., Yokoyama Y., Takekawa S. Combination chemoembolization therapy for hepatocellular carcinoma: mainly, using cisplatin (CDDP). Cancer Chemother. Pharmacol., 23 (Suppl.): S42-S44, 1989.
  39. Pelletier G., Roche A., Ink O., Anciaux M. L., Derhy S., Rougier P., Lenoir C., Attali P., Etienne J. P. A randomized trial of hepatic arterial chemoembolization in patients with unresectable hepatocellular carcinoma. J. Hepatol., 11: 181-184, 1990.[Medline]
  40. Group D’Etude et de Traitement du Carcinome Hepatocellulaire. A comparison of lipiodol chemoembolization and conservative treatment for unresectable hepatocellular carcinoma—Groupe d’Etude et de Traitement du Carcinome Hepatocellulaire. N. Engl. J. Med., 332: 1256-1261, 1995.
  41. Perry L. J., Stuart K., Stokes K. R., Clouse M. E. Hepatic arterial chemoembolization for metastatic neuroendocrine tumors. Surgery, 116: 1111-1116, 1994.[Medline]
  42. Mavligit G. M., Zukiwski A. A., Salem P. A., Lamki L., Wallace S. Regression of hepatic metastases from gastrointestinal leiomyosarcoma after hepatic arterial chemoembolization. Cancer (Phila.), 68: 321-323, 1991.[Medline]
  43. Leichman C. G., Fleming T. R., Muggia F. M., Tangen C. M., Ardalan B., Doroshow J. H., Meyers F. J., Holcombe R. F., Weiss G. R., Managlik A., Macdonald J. S. Phase II study of fluorouracil and its modulation in advanced colorectal cancer: a Southwest Oncology Group Study. J. Clin. Oncol., 13: 1303-1311, 1995.[Abstract]
  44. MetaAnalysis Group in Cancer. Efficacy of intravenous continuous infusion of 5-fluorouracil compared with bolus administration in advanced colorectal cancer. J. Clin. Oncol., 16: 301-308, 1998.[Abstract/Free Full Text]
  45. Leichman C. G. Prolonged infusion of fluorinated pyrimidines in gastrointestinal malignancies: a review of recent clinical trials. Cancer Invest., 12: 166-175, 1994.[Medline]
  46. Hansen R. M., Ryan L., Anderson T., Krzywda B., Quebbeman E., Benson A. B., III, Haller D. G., Tormey D. C. Phase III study of bolus versus infusion fluorouracil with or without cisplatin in advanced colorectal cancer. J. Natl. Cancer Inst., 88: 668-674, 1996.[Abstract/Free Full Text]
  47. Chang A. E., Schneider P. D., Sugarbaker P. H., Simpson C., Culnane M., Steinberg S. M. A prospective randomized trial of regional versus systemic continuous 5-fluorodeoxyuridine chemotherapy in the treatment of colorectal liver metastases. Ann. Surg., 206: 685-693, 1987.[Medline]
  48. Hohn D. C., Stagg R. J., Friedman M. A., Hannigan J. F., Rayner A., Ignoffo R. J., Acord P., Lewis B. J. 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. Oncol., 7: 1646-1654, 1989.[Abstract]
  49. Kemeny N., Daly J., Reichman B., Geller N., Botet J., Oderman P. Intrahepatic or systemic infusion of fluorodeoxyuridine in patients with liver metastases from colorectal carcinoma: a randomized trial. Ann. Intern. Med., 107: 459-465, 1987.
  50. Rougier P., Laplanche A., Huguier M., Hay J. M., Ollivier J. M., Escat J., Salmon R., Julien M., Roullet Audy J. C., Gallot D. Hepatic arterial infusion of floxuridine in patients with liver metastases from colorectal carcinoma: long-term results of a prospective randomized trial. J. Clin. Oncol., 10: 1112-1118, 1992.[Abstract]
  51. Martin J. K., Jr., O’Connell M. J., Wieand H. S., Fitzgibbons R. J., Jr., Mailliard J. A., Rubin J., Nagorney D. M., Tschetter L. K., Krook J. E. Intra-arterial floxuridine vs. systemic fluorouracil for hepatic metastases from colorectal cancer: a randomized trial. Arch. Surg., 125: 1022-1027, 1990.[Abstract]
  52. Meta-Analysis Group in Cancer. Reappraisal of hepatic arterial infusion in the treatment of nonresectable liver metastases from colorectal cancer. J. Natl. Cancer Inst., 88: 252-258, 1996.[Abstract/Free Full Text]
  53. Harmantas A., Rotstein L. E., Langer B. Regional versus systemic chemotherapy in the treatment of colorectal carcinoma metastatic to the liver. Is there a survival difference? Meta-analysis of the published literature. Cancer (Phila.), 78: 1639-1645, 1996.[Medline]
  54. Coldwell D. M., Mortimer J. E. Hepatic artery embolization in the treatment of hepatic malignancies. Regional Cancer Treatment, 3: 298-301, 1991.
  55. Lokich J. J., Ahlgren J. D., Cantrell J., Heim W. J., Wampler G. L., Gullo J. J., Fryer J. G., Alt D. E. A prospective randomized comparison of protracted infusional 5-fluorouracil with or without weekly bolus cisplatin in metastatic colorectal carcinoma: a Mid-Atlantic Oncology Program study. Cancer (Phila.), 67: 14-19, 1991.[Medline]
  56. Gori A., Conte P. F., Anfossi A. Neurotoxicity of intra-arterial cis-platinum (CDDP) in the locoregional treatment of diffuse liver metastasis of carcinoma of the colon and rectum. Minerva Chir., 45: 663-668, 1990.[Medline]
  57. Kelsen D. P., Hoffman J., Alcock N., Cheng E., Bailey E., Young C., Golbey R., Fortner J. Pharmacokinetics of cisplatin regional hepatic infusions. Am. J. Clin. Oncol., 5: 173-178, 1982.[Medline]
  58. Patt Y. Z., Boddie A. W., Charnsangavej C., Ajani J. A., Wallace S., Soski M., Claghorn L., Mavligit G. Hepatic arterial infusion with floxuridine and cisplatin: over-riding importance of antitumor effect versus degree of tumor burden as determinants of survival among patients with colorectal cancer. J. Clin. Oncol., 4: 1356-1364, 1986.[Abstract/Free Full Text]
  59. Wanebo H. J., Levy A. Alternating hepatic artery infusion of 5-fluorodeoxyuridine (FUdR), dexamethasone (dex) and leucovorin with systemic continuous infusion (CI) of 5-fluorouracil is associated with apparent survival benefit (median 18 months) in patients with unresectable liver metastases from colorectal cancer. Proc. Am. Soc. Clin. Oncol., 17: 271 1998.
  60. Coldwell D. M., Loper K. A. Regional anesthesia for hepat