Chemoembolization for Hepatocellular Carcinoma By Alan P. Venook, Robert J. Stagg, Brian J. Lewis, Judy L.Chase, Ernest J. Ring, Timothy P. Maroney, and David C. Hohn Fifty-one patients with unresectable hepatocellular carcinoma (HCC) were treated with Gelfoam (absorbable gelatin sterile powder; The Upjohn Co, Kalamazoo, MI) chemoembolization. A mixture of Gelfoam powder, contrast media, and three drugs (doxorubicin, mitomycin, and cisplatin) was injected under fluoroscopic guidance via a percutaneous catheter into the hepatic artery until stagnation of blood flow was achieved. Of the 51 patients, 50 are assessable for response, and all are assessable for toxicity and complications. The median percent of liver replacement was 50% (range, 15% to 95%). By conventional response criteria, there were 12 partial responses (PRs) (24%), 13 minor responses (MRs) (26%), 12 stabilization of disease (SD) (24%), and 13 (26%) progressive disease (PD). Tumor liquefaction was noted on computed tomographic (CT) scan in 35 of 50 patients (70%). Of the 34 patients with elevated alpha-
fetoprotein (AFP), 23 (68%) had a greater than 50% reduction following treatment. Responding patients were re-treated at the time of tumor progression if they still met the entry criteria. The median survival of assessable patients from the time of treatment was 207 days and from the diagnosis of the primary was 302 days. Fourteen patients remain alive at 3 months to 3 years following treatment. The vast majority of patients had transient pain, fever, nausea, and elevation in liver enzymes. Ascites developed in 14 patients. There were two treatment-related deaths: one from tumor hemorrhage and one from liver failure. Chemoembolization appears to have significant activity in patients with hepatocellular carcinoma and is relatively well tolerated. J Clin Oncol 8:1108- 1114. © 1990 by American Society of Clinical Oncology.
HEPATOCELLULAR
patient with otherwise unresectable nonmetastatic fibrolamellar HCC.3 For unresectable HCC, the most widely used treatment has been systemic chemotherapy. Doxorubicin, fluorouracil, and mitomycin, used singly or in combination, produce response rates of about 20% and do not impact on survival.4 Intraarterial (IA) chemotherapy with floxuridine or doxorubicin has produced response rates of approximately 50%.5,6 However, this approach requires either a laparotomy for surgical placement of a catheter or a prolonged hospitalization for drug administration via a percutaneous transarterial catheter. Both are technically cumbersome, and the latter is associated with significant complications. External beam radiotherapy is
carcinoma (HCC) is one of the most prevalent malignancies in the world. However, in the United States HCC is
relatively rare, with only about 3,000 to 4,000 cases diagnosed each year. Untreated HCC car-
ries an extremely poor prognosis, with a median survival of approximately 3 to 6 months depending upon tumor extent, the presence or absence of
cirrhosis, and performance status.' Several modalities have been used in the treat-
ment of patients with HCC. Surgical resection is potentially curative, but is only feasible in that small minority of patients with solitary or unilobar hepatic lesions without metastases who do
not have significant underlying liver dysfunction. 2 Total hepatectomy with orthotopic liver transplantation is of limited benefit except in the rare
From the Cancer Research Institute, the Department of Radiology, and the Division of Clinical Pharmacy, University of California,San Francisco,CA; and the Departmentof Surgery, MD Anderson Cancer Center, Houston, TX. Submitted July 10, 1989; accepted January 30, 1990. Address reprintrequests to Alan P. Venook, MD, Cancer Research Institute,M-1282/Box 0128, University of California Medical Center,San Francisco,CA 94143. © 1990 by American Society of Clinical Oncology. 0732-183X/90/0806-0016$3.00/0
1108
limited by the radiosensitivity of normal hepatocytes. Radioactive isotope of iodine 131 (131I)-
labeled antiferritin antibodies, in conjunction with intravenous (IV) chemotherapy and external beam irradiation, appear to have some activity in alpha-fetoprotein (AFP)-negative patients,7 's although the role for this modality is yet to be determined. Because liver tumors derive the vast majority of their blood flow from the hepatic artery while normal liver derives its blood supply from the portal vein, 9 selective tumor ischemia may be
Journalof Clinical Oncology, Vol 8, No 6 (June), 1990: pp 1108-1114
Downloaded from ascopubs.org by McMaster University on September 14, 2019 from 130.113.111.210 Copyright © 2019 American Society of Clinical Oncology. All rights reserved.
CHEMOEMBOLIZATION FOR HEPATOCELLULAR CARCINOMA
induced by ligation of the hepatic artery.'o However, this is generally of only transient benefit," presumably because of the rapid development of collateral arterial blood flow to the tumor.12 Percutaneous transarterial embolization of the hepatic artery with vasoocclusive materials has been done safely with some efficacy in patients with primary and metastatic liver tumors.'316 The use of degradable vasoocclusive particles theoretically minimizes the collateralization of blood flow by creating a more distal vascular blockade and allows for re-treatment. The term chemoembolization has variously been used to describe the combination of transarterial embolization with IV or IA chemotherapy delivered sequentially or concomitantly. Because of the variation in chemoembolization protocols, it is difficult to compare the relative efficacy of this modality to other treatments, although encouraging response rates have been reported both in the palliation of unresectable tumors and as a means of cytoreduction and devascularization before surgery.'7- 24 Some efficacy has also been
demonstrated for chemoembolization in the treatment of metastatic liver tumors.2 5 2 8
This report describes the results of a phase I-II trial of Gelfoam (absorbable gelatin sterile powder; the Upjohn Co, Kalamazoo, MI) chemoembolization in patients with unresectable, liverpredominant HCC. PATIENTS AND METHODS Beginning in February 1986, patients with unresectable, biopsy-proven, liver-dominant HCC were considered for entry into the study. Entry criteria included a Karnofsky performance status _ 60, a prothrombin time less than 15 seconds, serum creatinine less than 1.6 mg/dL, WBC count greater than 3 ,000/pL, and a platelet count greater than 60,000/uL. Patients with massive ascites, bilirubin greater than 7 mg/dL, or encephalopathy were excluded, as were patients with complete portal vein occlusion. All patients had either an abdominal computed tomographic (CT) scan or magnetic resonance imaging (MRI) within 2 weeks before treatment. A complete blood and platelet count, liver function tests, albumin, coagulation parameters, BUN, creatinine, and AFP were obtained within 1week before treatment. Informed written consent was obtained from all patients. Before treatment, patients were hydrated for a minimum of 4 hours at 150 to 250 cc/h to establish urine output of at least 100 cc/h. The last 31 patients received prophylactic IV metronidazole after two patients being treated on a parallel protocol for metastatic liver tumors developed hepatic abscesses caused by clostridial organisms. Under local anesthesia and mild systemic sedation, the femoral artery was catheterized using the Seldinger tech-
1109
nique. The portal vein was evaluated by arterial portography with injections into either the splenic or superior mesenteric artery. Patients found to have occluded portal veins were not treated. Those with patent portal veins underwent selective catheterization of the common hepatic artery to identify the precise positions of the gastroduodenal, right gastric, and right and left hepatic arteries. Gastric branches were scrupulously avoided to prevent gastroduodenal misperfusion, but no special precautions were taken to avoid embolization of the cystic artery. Patients having minimal tumor burden were treated with superselective embolization of only those vessels supplying the tumor. Patients with underlying cirrhosis or hepatitis, extensive bilobar disease, massive lesions, or extremely vascular tumors were treated with staged unilobar chemoembolizations at 4- to 6-week intervals in order to preserve adequate liver function. The remaining patients received whole-liver chemoembolization. When anatomic variations were encountered, branches were individually chemoembolized. The chemoembolization mixture contained Gelfoam powder (a 40- to 50-micron microfibrillan collagen) at a concentration of 25 to 30 mg/cc, suspended in a total volume of 10 cc of radiopaque contrast media and chemotherapy. The drugs and doses used were mitomycin 10 mg, cisplatin 50 mg, and doxorubicin 50 mg. The doxorubicin dose was reduced to 20 mg in those patients with a bilirubin between 3 to 7 mg/dL, WBC count less than 3.0, or platelet count less than 90,000. Doxorubicin was excluded in patients with a prior total systemic dose greater than 450 mg/m 2 and an abnormal multigated angiogram or heart disease. This suspension was infused under fluoroscopic control until stagnation of hepatic arterial blood flow was achieved. If blood flow remained antegrade following the administration of 10 cc of this mixture, bland Gelfoam powder or pledgets were infused until the complete cessation of flow was achieved. Following treatment, vigorous hydration and metronidazole were continued for at least 48 hours. Postprocedure pain was managed with parenteral analgesics, and nausea and vomiting were treated with antiemetics and/or nasogastric suction. Fevers were evaluated as befitted the clinical situation. Liver function tests, BUN, and creatinine were monitored daily while the patient was in the hospital. Responses were judged by CT or MRI scanning according to the following definitions: complete response (CR), the complete regression of all hepatic tumor; partial response (PR), greater than 50% decrease in the sum of the products of perpendicular diameters of the hepatic lesions; minor response (MR), 25% to 50% reduction in the sum of the products of perpendicular diameters of all hepatic lesions; stabilization of disease (SD), no detectable change in hepatic tumor size; progressive disease (PD), the appearance of new hepatic lesions or a greater than 25% increase in the sum of the products of perpendicular diameters of all hepatic lesions. In addition to these standard response criteria, the presence of radiographic changes consistent with liquefaction necrosis were noted. These changes included a decrease in tumor density or enhancement, clear demarcation of tumor margins, and/or the appearance of air-fluid levels within the lesions. Scans were repeated every 2 months and patients who responded were offered repeat treatment at the time of liver progression if they still met the original entry criteria. Liver
Downloaded from ascopubs.org by McMaster University on September 14, 2019 from 130.113.111.210 Copyright © 2019 American Society of Clinical Oncology. All rights reserved.
VENOOK ET AL
1110 function tests and AFP were followed serially. Survival was recorded both from the time of initial presentation as well as the time of study entry. Toxicity was scored following standard grades.
RESULTS
The characteristics of the 51 patients enrolled in the study are listed in Table 1. One patient was lost to follow-up and is excluded from analysis. Six patients could not be evaluated for response (treatment-related death [two patients], nontreatment-related early death [four patients]) but are considered treatment failures for statistical purposes. All but one patient were ambulatory and the median percentage of liver replaced by tumor was 50% (range, 15% to 95%) as estimated by CT scan. Twelve patients had minor extrahepatic disease at presentation and 16 had received prior therapy (IV or IA chemotherapy, radiotherapy, surgery, radioimmunotherapy). Twenty-nine patients (57%) had chronic hepatitis B surface antigenemia and/or biopsy-proven cirrhosis. The majority of patients (34 of 51) had tumors that produced AFP. As indicated in Table 2, 12 of 50 (24%) patients had PRs (Fig 1) and 13 patients (26%) had MRs. Responses were seen in patients who received selective, as well as whole liver treatment. Twelve patients (24%) had SD and 13 patients (26%) had either progressive intrahepatic disease at their initial posttreatment evaluation or were deemed treatment failures by the above criteria. Liquefaction necrosis (Fig 2) was observed on CT scan in 35 of 50 (70%) patients. Of those patients whose tumor made AFP, 23 of 34 (68%) had a greater than 50% reduction in serum level after treatment. The median days of hospitalization for the procedure was 4 days (range, 3 to 15 days) and the median interval between treatments was 4 months (range, 2 to Table 1. Patient Characteristics (N = 51) Median age Male:female Median % liver replacement Extrahepatic disease Prior therapy Cirrhosis documented Hepatitis BsAg + Minimal ascites AFP producing
57 (16-77) 33:18 50% (15%-95%) 12 16 19 21 5 34
Table 2. CT Response of Hepatoma to Chemoembolization (N = 50)
Complete response Partial response Minor response Stable disease Progression/early death
Without
With
Necrosis
Necrosis
Total
(%)
0 0 1 2 12
0 12 12 10 1
0(0) 12 (24) 13(26) 12 (24) 13 (26)
24+). The 50 patients received a total of 72 treatments (range, one to five). Re-treatment was not offered when there was extrahepatic progression (14 patients), failure to respond in the liver (12 patients), combined hepatic and major extrahepatic progression (two patients), loss of follow-up (two patients), and technical limitations (one patient). Ten patients remain eligible for further therapy. The median survival of all assessable patients from the time of diagnosis of the primary tumor was 302 days (range, 0 to 70+ months) and the median survival from the institution of therapy was 207 days (range, 0 to 37+ months) (Fig 3). Presently, 14 patients remain alive at 3 to 37+ months after their first chemoembolization. Every patient had transient increases in liver function tests, particularly in lactic dehydrogenase. Almost every patient experienced some degree of nausea and/or vomiting, pain, and fever (Table 3). Peripheral edema developed in 14 patients and was accompanied by the development or exacerbation of ascites in 12 patients. In all but three patients, these symptoms resolved. Transient encephalopathy developed in three patients, and three had increases of creatinine greater than twofold, one of whom required temporary hemodialysis. No patients had significant hematologic toxicity. Six patients had significant complications (Table 4), two of which resulted in death; one exsanguinated from his tumor, and the other died from liver failure. DISCUSSION
The therapy of HCC has been extremely disappointing. Systemic chemotherapy produces a response rate of no more than 20% and does not impact favorably on the median survival of 3 to 6 months for the average patient with HCC. 4 Hepatic artery infusional chemotherapy may be more active, but has failed to improve survivor-
Downloaded from ascopubs.org by McMaster University on September 14, 2019 from 130.113.111.210 Copyright © 2019 American Society of Clinical Oncology. All rights reserved.
CHEMOEMBOLIZATION FOR HEPATOCELLULAR CARCINOMA
Fig 1.
1111
(A) Pretreatment CT scan and (B) CT scan 8 months after staged chemoembolization of a patient with HCC,
representing a partial response.
ships'6 and is associated with significant toxicity, particularly in patients with underlying liver disease. 29 Hepatic dearterialization accomplished by the ligation of the hepatic artery has been attempted in patients with HCC. No discernible improvement was seen in survival, and it has an operative mortality rate of as high as 23%.30 One-shot hepatic artery occlusion with particles alone has induced clinical responses in some patients, but the actual activity is difficult to measure due to the lack of rigorous response criteria in some trials.14,20 '22,31 Our pilot experience with 17 patients with a variety of hepatic neoplasms who were treated with particle embolization alone yielded little evidence of tumor necrosis, no change in liver function tests, and no pain following treatment. Although polyvinyl alcohol embolization without chemotherapy is active against
islet cell tumors, 16 it is our impression that the addition of chemotherapy was necessary in the treatment of HCC. The majority of the experience combining chemotherapy and particle embolization to treat hepatic tumors comes from Asia, where the high incidence of primary liver cancer has led to vigorous investigation of novel treatment modalities. In our trial, Gelfoam powder (diameter of 40 to 50 gm) was used as the embolizing material with the expectation that it would produce a more distal arteriolar/capillary occlusion with less collateralization to the tumor than would occur with larger Gelfoam particles or unprocessed polyvinyl alcohol (250 to 590 am). Gelfoam undergoes degradation and resorption in the circulation, and the vessels recanalize within 48 to 72 hours,32 while undegradable polyvinyl alcohol causes permanent vascular occlusion. 33
Fig 2. (A) Pretreatment CT scan and (B)CT scan 8 months after whole-liver chemoembolization for HCC. This demonstrates the appearance of liquefaction necrosis without evidence of tumor shrinkage.
Downloaded from ascopubs.org by McMaster University on September 14, 2019 from 130.113.111.210 Copyright © 2019 American Society of Clinical Oncology. All rights reserved.
1112
VENOOK ET AL Table 4. Complications of Chemoembolization (N = 51)
1
Type of Complication
No. of Patients
Tumor hemorrhage
1
Tumor rupture
1
Liver failure Cerebral hemorrhage Liver abscess Tumor lysis syndrome
1 1 1 1
(j~ 4-,
0.
U
Time (months) Fig 3. Actuaria I survival of all assessable patients with HCC from the time of initiation of chemoembolization.
This difference allows for repeated treatments of the same vascu lar bed with Gelfoam, whereas with polyvinyl a alcohol, flow is permanently redistributed into col lateral vessels. The most ac tive chemotherapeutic agents in HCC are fluoro uracil, doxorubicin, and mitomycin. Meaningful doses of fluorouracil could not be incorporated into the chemoembolization mixture due to the low drug concentration of the available soluti )n. Cisplatin was combined with doxorubicin and mitomycin based on the unpublished pilot expterience of Aigner et al (personal communication. February 1986), who found the treatment to be active and well-tolerated in a few Table 3. Toxic ity of Chemoembolization (N Type of To'Xicity xic, Transient LFT ele vation Pain Nausea and/or vomiting Fever Edema Ascites Transient encephhalopathy Alopecia Transient renal d ysfunction* Infection Hematologic
=
51)
patients. Since cisplatin, mitomycin, and doxorubicin have relatively slow total body clearances and limited hepatic extractions, they are less than ideal for unmodified regional infusion. 34 However, enhanced regional drug exposure may be achieved when IA administration of these drugs is combined with the interruption of arterial blood flow, as has been demonstrated with chemoembolization of coloreccisplatin/collagen tal metastases 27 and with mitomycin when admin35 istered with degradable starch microspheres. While the radiographically documented PR rate observed in this trial is only 24%, the
appearance of liquefaction on CT scan, as occurred in 70% of our patients, has been shown to correlate histopathologically with significant tumor necrosis.'5, 24 This has been confirmed in four of our patients who went to surgery following chemoembolization and were found to have mostly necrotic tumor. In other words, liquefaction necrosis may represent a clinically significant antitumor effect (Figs 1 and 2). Therefore, applying standard radiographic response criteria to judge this type of treatment may underestimate its activity. We feel that this treatment had a discernible and favorable impact on our patients. Since we serve as a referral center, our patients tend to have advanced disease (a median of 50% liver replacement). Furthermore, a majority of our
"LI 'r"'"
T
"rr c"rr "r'-' 'Y lllll·P 51nlllrPn No. of (%positive and/or had evidence of cirrhosis, charac51 (100) teristics that portend a bad outcome regardless of 48 (94) the treatment rendered. 29 The median survival 47 (92) for patients with HCC and cirrhosis may be as 46 (90) 14(27) little as 1.2 months.' Doci et a129 reported a jJIAIU
V
i T.' Lfh
f
I
i~ L LW
AS04
11
1L1
No. of Patients Patients
12(24)
median survival of 3.5 months in a group of
3 (6) 3 (6)
patients similar to ours who were treated with IA
3(6)
chemotherapy. Therefore, a median survival of 7
(2) 0(0)
months from first treatment, with 25% of the patients surviving more than 20 months from the time of treatment, seems significant. The largest experience with chemoemboliza-
1
Abbreviation: LFT, liver function test. *Twofold or greatetr increase in serum creatinine.
Downloaded from ascopubs.org by McMaster University on September 14, 2019 from 130.113.111.210 Copyright © 2019 American Society of Clinical Oncology. All rights reserved.
CHEMOEMBOLIZATION FOR HEPATOCELLULAR CARCINOMA
tion for HCC comes from Asia.15,18-2 4 However, differences in patient selection and methodology, including choice of drugs and particle, the route of drug delivery, the inclusion of Lipiodol UltraFluide (Guerbet, Aulnay sous Bois, France) in the mixture, 23 and the sequence in which these are administered, make it difficult to compare those results with ours. In Japan, chemoembolization is also used preoperatively in patients with resectable tumors in an attempt to decrease tumor vascularity and the risk of hematogenous metastases at the time of liver mobilization. No benefit for this approach has been shown.36 Toxicity was significant but predictable and mostly self-limited. In general, patients who received whole liver chemoembolization had more pronounced toxicity than those treated with selective procedures. Although virtually every patient experienced nausea, vomiting, pain, and fever, the majority were ambulatory within 4 days and largely asymptomatic within 2 weeks. Patients' quality of life was further aided by the absence of repeated chemotherapy treatments until tumor progression became evident. The median hospital stay of 4 days is less than that for a single standard cycle of IA therapy. Although Takayasu et a137 reported a 50% incidence of postembolization cholecystitis due to occlusion of the cystic artery, we encountered no clinically apparent gall bladder complications. Since we used no
1113
special precautions to avoid embolizing the cystic artery, this discrepancy with Takayasu et al's experience is unexplained. The two treatmentrelated deaths occurred in patients with very little hepatic reserve (one with extensive cirrhosis and one with a large, necrotic tumor). Our experience in this trial leads us to conclude that patients with HCC who are good candidates for chemoembolization are those who have unresectable, liver-predominant disease, a Karnofsky performance status of 60% or better, no ascites, and adequate hepatic reserve (albumin > 2.5 g/dL, bilirubin < 3 mg/dL, normal prothrombin time, and partial thromboplastin time). While this trial demonstrates encouraging activity of chemoembolization in HCC, the following questions regarding this modality remain unresolved: (1) Should re-treatment be used at set intervals rather than at the time of hepatic progression, as used in this trial? (2) Which of these or other drugs should be used in the chemoembolization mixture? (3) Is the selection of the specific embolization material important for a given tumor type? (4) How does one reproducibly and systematically describe the effects of this treatment given the experience that changes in tumor volume may not tell the entire story? While these questions are being addressed, we believe the treatment should remain confined to protocols at dedicated treatment centers.
REFERENCES 1. Smalley SR, Moertel CG, Hilton JF, et al: Hepatoma in the noncirrhotic liver. Cancer 62:1414-1424, 1988 2. Malt RA: Surgery for hepatic neoplasms. N Engl J Med 313:1591-1596, 1985 3. Starzl TE, Iwatsuki S, Shaw BW, et al: Treatment of fibrolamellar hepatoma with partial hepatectomy and transplantation of the liver. Surg Gynecol Obstet 162:145-148, 1986 4. Lewis BJ, Friedman MA: Current status of chemotherapy for hepatoma, in Ogawa M (ed): Chemotherapy of Hepatic Tumors. Princeton, NJ Excerpta Medica, 1984, pp 63-74 5. Wellwood JM, Cady B, Oberfield RA: Treatment of primary liver cancer: Response with regional chemotherapy. J Clin Oncol 5:25-31, 1979 6. Urist MM, Balch CM: Intraarterial chemotherapy for hepatoma using Adriamycin administered via the implantable infusion pump. Proc Am Soc Clin Oncol 3:146, 1983 (abstr) 7. Order SE, Stillwagon GB, Klein JL, et al: Iodine 131 antiferritin, a new treatment modality in hepatoma: A Radiation Therapy Oncology Group study. J Clin Oncol 3:1573-1582, 1985
8. Order S, Pajak T, Klein J, et al: A randomized prospective trial in nonresectable hepatoma comparing Adriamycin and 5-fluorouracil + 131-I antiferritin: An RTOG study. Proc Am Soc Clin Oncol 8:99, 1989 (abstr) 9. Breedis C, Young G: The blood supply of neoplasms in the liver. Am J Pathol 30:969-985, 1954 10. Markowitz J: The hepatic artery. Surg Gynecol Obstet 95:644-646, 1952 11. Sparks F, Mosher M, Hallauer W, et al: Hepatic artery ligation and postoperative chemotherapy for hepatic metastases: Clinical and pathophysiological results. Cancer 35:1074-1082, 1975 12. Charnsangavej C, Chuang VP, Wallace S, et al: Angiographic classification of hepatic arterial collaterals. Radiology 144:485-494, 1982 13. Goldstein HM, Wallace S, Anderson JH, et al: Transcatheter occlusion of abdominal tumors. Radiology 120:539-545, 1976 14. Clouse ME, Lee RGL, Duszlak EJ: Peripheral hepatic artery embolization for primary and secondary hepatic neoplasms. Radiology 147:407-411, 1983 15. Yamada R, Sato M, Kawabata M, et al: Hepatic
Downloaded from ascopubs.org by McMaster University on September 14, 2019 from 130.113.111.210 Copyright © 2019 American Society of Clinical Oncology. All rights reserved.
1114
VENOOK ET AL
artery embolization in 120 patients with unresectable hepatoma. Radiology 148:397-401, 1983 16. Ajani JA, Carrasco H, Charnsangavej C, et al: Islet cell tumors metastatic to the liver: Effective palliation by sequential hepatic artery embolization. Ann Intern Med 108:340-344, 1988 17. Fujimoto S, Miyazaki M, Endoh F, et al: Biodegradable mitomycin c microspheres given intra-arterial for inoperable hepatic cancer. Cancer 56:2404-2410, 1985 18. Ohnishi K, Tsuchiya S, Nakayama T, et al: Arterial chemoembolization of hepatocellular carcinoma with mitomycin c microcapsules. Radiology 152:51-55, 1984 19. Sakurai M, Okamura J, Kuroda C: Transcatheter chemoembolization effective for treating hepatocellular carcinoma. Cancer 54:387-392, 1984 20. Lin D, Liaw Y, Lee T, et al: Hepatic arterial embolization in patients with unresectable hepatocellular carcinoma-A randomized controlled trial. Gastroenterology 94:453-456, 1988 21. Soga K, Nomoto M, Ichida T, et al: Clinical evaluation of transcatheter arterial embolization and one-shot chemotherapy in hepatocellular carcinoma. Hepatogastroenterology 35:116-120, 1988 22. Sato Y, Fujiwara K, Ogata I, et al: Transcatheter arterial embolization for hepatocellular carcinoma. Cancer 55:2822-2825, 1985 23. Shimamura Y, Gunven P, Takenaka Y, et al: Combined peripheral and central chemoembolization of liver tumors. Cancer 61:238-242, 1988 24. Takayasu K, Moriyama N, Muramatsu Y, et al: Hepatic arterial embolization for hepatocellular carcinoma. Radiology 150:661-665, 1984 25. Carrasco C, Wallace S, Charnsangavej C, et al: Treatment of hepatic metastases in ocular melanoma. JAMA 255:3152-3154, 1986 26. Daniels J, Daniels A, Quinn M, et al: Phase I trial with cisplatin or mitomycin hepatic chemoembolization with Angiostat collagen for embolization in patients with colorectal cancer. Proc Am Soc Clin Oncol 7:385, 1988 (abstr)
27. Modiano M, Holdsworth M, Alberts D, et al: Control of hepatic metastases from colorectal carcinoma with selective intraarterial chemoembolization with Angiostat collagen, CDDP, mitomycin and Adriamycin. The University of Arizona experience. Proc Am Soc Clin Oncol 8:105, 1989 (abstr) 28. Venook A, Chase J, Stagg R, et al: A phase I-II trial of angiostat/mitomycin-c hepatic chemoembolization for colorectal cancer metastatic to the liver. Proc Am Soc Clin Oncol 8:120, 1989 (abstr) 29. Doci R, Bignami P, Bozzetti F, et al: Intrahepatic chemotherapy for unresectable hepatocellular carcinoma. Cancer 61:1983-1987, 1988 30. Almersjo O, Bengmark S, Rudenstam CM, et al: Evaluation of hepatic dearterialization in primary and secondary cancer of the liver. Am J Surg 124:5-9, 1972 31. Allison DJ, Jordan H, Hennessy O: Therapeutic embolisation of the hepatic artery: A review of 75 procedures. Lancet 1:595-599, 1985 32. Barth KH, Strandberg JD, White RI: Long term follow-up of transcatheter embolization with autologous clot, oxycel and Gelfoam in domestic swine. Invest Radiol 12:273280, 1977 33. Chuang VP, Wallace S, Soo CS, et al: Therapeutic ivalon embolization of hepatic tumors. Am J Roent 138:289294, 1982 34. Collins JM: Pharmacologic rationale for regional drug delivery. J Clin Oncol 2:498-504, 1984 35. Andersson M, Aronsen KF, Balch C, et al: Pharmacokinetics of intra-arterial mitomycin C with or without degradable starch microspheres (DSM) in the treatment of nonresectable liver cancer. Acta Oncol 28:219-222, 1989 36. Nagasue N, Galizia G, Kohno H, et al: Adverse effects of preoperative hepatic artery chemoembolization for resectable hepatocellular carcinoma: A retrospective comparison of 138 liver resections. Surgery 106:81-86, 1989 37. Takayasu K, Moriyama N, Muramatsu Y, et al: Gallbladder infarction after hepatic artery embolization. AJR 144:135-138, 1985
Downloaded from ascopubs.org by McMaster University on September 14, 2019 from 130.113.111.210 Copyright © 2019 American Society of Clinical Oncology. All rights reserved.
fetoprotein (AFP), 23 (68%) had a greater than 50% reduction following treatment. Responding patients were re-treated at the time of tumor progression if they still met the entry criteria. The median survival of assessable patients from the time of treatment was 207 days and from the diagnosis of the primary was 302 days. Fourteen patients remain alive at 3 months to 3 years following treatment. The vast majority of patients had transient pain, fever, nausea, and elevation in liver enzymes. Ascites developed in 14 patients. There were two treatment-related deaths: one from tumor hemorrhage and one from liver failure. Chemoembolization appears to have significant activity in patients with hepatocellular carcinoma and is relatively well tolerated. J Clin Oncol 8:1108- 1114. © 1990 by American Society of Clinical Oncology.
HEPATOCELLULAR
patient with otherwise unresectable nonmetastatic fibrolamellar HCC.3 For unresectable HCC, the most widely used treatment has been systemic chemotherapy. Doxorubicin, fluorouracil, and mitomycin, used singly or in combination, produce response rates of about 20% and do not impact on survival.4 Intraarterial (IA) chemotherapy with floxuridine or doxorubicin has produced response rates of approximately 50%.5,6 However, this approach requires either a laparotomy for surgical placement of a catheter or a prolonged hospitalization for drug administration via a percutaneous transarterial catheter. Both are technically cumbersome, and the latter is associated with significant complications. External beam radiotherapy is
carcinoma (HCC) is one of the most prevalent malignancies in the world. However, in the United States HCC is
relatively rare, with only about 3,000 to 4,000 cases diagnosed each year. Untreated HCC car-
ries an extremely poor prognosis, with a median survival of approximately 3 to 6 months depending upon tumor extent, the presence or absence of
cirrhosis, and performance status.' Several modalities have been used in the treat-
ment of patients with HCC. Surgical resection is potentially curative, but is only feasible in that small minority of patients with solitary or unilobar hepatic lesions without metastases who do
not have significant underlying liver dysfunction. 2 Total hepatectomy with orthotopic liver transplantation is of limited benefit except in the rare
From the Cancer Research Institute, the Department of Radiology, and the Division of Clinical Pharmacy, University of California,San Francisco,CA; and the Departmentof Surgery, MD Anderson Cancer Center, Houston, TX. Submitted July 10, 1989; accepted January 30, 1990. Address reprintrequests to Alan P. Venook, MD, Cancer Research Institute,M-1282/Box 0128, University of California Medical Center,San Francisco,CA 94143. © 1990 by American Society of Clinical Oncology. 0732-183X/90/0806-0016$3.00/0
1108
limited by the radiosensitivity of normal hepatocytes. Radioactive isotope of iodine 131 (131I)-
labeled antiferritin antibodies, in conjunction with intravenous (IV) chemotherapy and external beam irradiation, appear to have some activity in alpha-fetoprotein (AFP)-negative patients,7 's although the role for this modality is yet to be determined. Because liver tumors derive the vast majority of their blood flow from the hepatic artery while normal liver derives its blood supply from the portal vein, 9 selective tumor ischemia may be
Journalof Clinical Oncology, Vol 8, No 6 (June), 1990: pp 1108-1114
Downloaded from ascopubs.org by McMaster University on September 14, 2019 from 130.113.111.210 Copyright © 2019 American Society of Clinical Oncology. All rights reserved.
CHEMOEMBOLIZATION FOR HEPATOCELLULAR CARCINOMA
induced by ligation of the hepatic artery.'o However, this is generally of only transient benefit," presumably because of the rapid development of collateral arterial blood flow to the tumor.12 Percutaneous transarterial embolization of the hepatic artery with vasoocclusive materials has been done safely with some efficacy in patients with primary and metastatic liver tumors.'316 The use of degradable vasoocclusive particles theoretically minimizes the collateralization of blood flow by creating a more distal vascular blockade and allows for re-treatment. The term chemoembolization has variously been used to describe the combination of transarterial embolization with IV or IA chemotherapy delivered sequentially or concomitantly. Because of the variation in chemoembolization protocols, it is difficult to compare the relative efficacy of this modality to other treatments, although encouraging response rates have been reported both in the palliation of unresectable tumors and as a means of cytoreduction and devascularization before surgery.'7- 24 Some efficacy has also been
demonstrated for chemoembolization in the treatment of metastatic liver tumors.2 5 2 8
This report describes the results of a phase I-II trial of Gelfoam (absorbable gelatin sterile powder; the Upjohn Co, Kalamazoo, MI) chemoembolization in patients with unresectable, liverpredominant HCC. PATIENTS AND METHODS Beginning in February 1986, patients with unresectable, biopsy-proven, liver-dominant HCC were considered for entry into the study. Entry criteria included a Karnofsky performance status _ 60, a prothrombin time less than 15 seconds, serum creatinine less than 1.6 mg/dL, WBC count greater than 3 ,000/pL, and a platelet count greater than 60,000/uL. Patients with massive ascites, bilirubin greater than 7 mg/dL, or encephalopathy were excluded, as were patients with complete portal vein occlusion. All patients had either an abdominal computed tomographic (CT) scan or magnetic resonance imaging (MRI) within 2 weeks before treatment. A complete blood and platelet count, liver function tests, albumin, coagulation parameters, BUN, creatinine, and AFP were obtained within 1week before treatment. Informed written consent was obtained from all patients. Before treatment, patients were hydrated for a minimum of 4 hours at 150 to 250 cc/h to establish urine output of at least 100 cc/h. The last 31 patients received prophylactic IV metronidazole after two patients being treated on a parallel protocol for metastatic liver tumors developed hepatic abscesses caused by clostridial organisms. Under local anesthesia and mild systemic sedation, the femoral artery was catheterized using the Seldinger tech-
1109
nique. The portal vein was evaluated by arterial portography with injections into either the splenic or superior mesenteric artery. Patients found to have occluded portal veins were not treated. Those with patent portal veins underwent selective catheterization of the common hepatic artery to identify the precise positions of the gastroduodenal, right gastric, and right and left hepatic arteries. Gastric branches were scrupulously avoided to prevent gastroduodenal misperfusion, but no special precautions were taken to avoid embolization of the cystic artery. Patients having minimal tumor burden were treated with superselective embolization of only those vessels supplying the tumor. Patients with underlying cirrhosis or hepatitis, extensive bilobar disease, massive lesions, or extremely vascular tumors were treated with staged unilobar chemoembolizations at 4- to 6-week intervals in order to preserve adequate liver function. The remaining patients received whole-liver chemoembolization. When anatomic variations were encountered, branches were individually chemoembolized. The chemoembolization mixture contained Gelfoam powder (a 40- to 50-micron microfibrillan collagen) at a concentration of 25 to 30 mg/cc, suspended in a total volume of 10 cc of radiopaque contrast media and chemotherapy. The drugs and doses used were mitomycin 10 mg, cisplatin 50 mg, and doxorubicin 50 mg. The doxorubicin dose was reduced to 20 mg in those patients with a bilirubin between 3 to 7 mg/dL, WBC count less than 3.0, or platelet count less than 90,000. Doxorubicin was excluded in patients with a prior total systemic dose greater than 450 mg/m 2 and an abnormal multigated angiogram or heart disease. This suspension was infused under fluoroscopic control until stagnation of hepatic arterial blood flow was achieved. If blood flow remained antegrade following the administration of 10 cc of this mixture, bland Gelfoam powder or pledgets were infused until the complete cessation of flow was achieved. Following treatment, vigorous hydration and metronidazole were continued for at least 48 hours. Postprocedure pain was managed with parenteral analgesics, and nausea and vomiting were treated with antiemetics and/or nasogastric suction. Fevers were evaluated as befitted the clinical situation. Liver function tests, BUN, and creatinine were monitored daily while the patient was in the hospital. Responses were judged by CT or MRI scanning according to the following definitions: complete response (CR), the complete regression of all hepatic tumor; partial response (PR), greater than 50% decrease in the sum of the products of perpendicular diameters of the hepatic lesions; minor response (MR), 25% to 50% reduction in the sum of the products of perpendicular diameters of all hepatic lesions; stabilization of disease (SD), no detectable change in hepatic tumor size; progressive disease (PD), the appearance of new hepatic lesions or a greater than 25% increase in the sum of the products of perpendicular diameters of all hepatic lesions. In addition to these standard response criteria, the presence of radiographic changes consistent with liquefaction necrosis were noted. These changes included a decrease in tumor density or enhancement, clear demarcation of tumor margins, and/or the appearance of air-fluid levels within the lesions. Scans were repeated every 2 months and patients who responded were offered repeat treatment at the time of liver progression if they still met the original entry criteria. Liver
Downloaded from ascopubs.org by McMaster University on September 14, 2019 from 130.113.111.210 Copyright © 2019 American Society of Clinical Oncology. All rights reserved.
VENOOK ET AL
1110 function tests and AFP were followed serially. Survival was recorded both from the time of initial presentation as well as the time of study entry. Toxicity was scored following standard grades.
RESULTS
The characteristics of the 51 patients enrolled in the study are listed in Table 1. One patient was lost to follow-up and is excluded from analysis. Six patients could not be evaluated for response (treatment-related death [two patients], nontreatment-related early death [four patients]) but are considered treatment failures for statistical purposes. All but one patient were ambulatory and the median percentage of liver replaced by tumor was 50% (range, 15% to 95%) as estimated by CT scan. Twelve patients had minor extrahepatic disease at presentation and 16 had received prior therapy (IV or IA chemotherapy, radiotherapy, surgery, radioimmunotherapy). Twenty-nine patients (57%) had chronic hepatitis B surface antigenemia and/or biopsy-proven cirrhosis. The majority of patients (34 of 51) had tumors that produced AFP. As indicated in Table 2, 12 of 50 (24%) patients had PRs (Fig 1) and 13 patients (26%) had MRs. Responses were seen in patients who received selective, as well as whole liver treatment. Twelve patients (24%) had SD and 13 patients (26%) had either progressive intrahepatic disease at their initial posttreatment evaluation or were deemed treatment failures by the above criteria. Liquefaction necrosis (Fig 2) was observed on CT scan in 35 of 50 (70%) patients. Of those patients whose tumor made AFP, 23 of 34 (68%) had a greater than 50% reduction in serum level after treatment. The median days of hospitalization for the procedure was 4 days (range, 3 to 15 days) and the median interval between treatments was 4 months (range, 2 to Table 1. Patient Characteristics (N = 51) Median age Male:female Median % liver replacement Extrahepatic disease Prior therapy Cirrhosis documented Hepatitis BsAg + Minimal ascites AFP producing
57 (16-77) 33:18 50% (15%-95%) 12 16 19 21 5 34
Table 2. CT Response of Hepatoma to Chemoembolization (N = 50)
Complete response Partial response Minor response Stable disease Progression/early death
Without
With
Necrosis
Necrosis
Total
(%)
0 0 1 2 12
0 12 12 10 1
0(0) 12 (24) 13(26) 12 (24) 13 (26)
24+). The 50 patients received a total of 72 treatments (range, one to five). Re-treatment was not offered when there was extrahepatic progression (14 patients), failure to respond in the liver (12 patients), combined hepatic and major extrahepatic progression (two patients), loss of follow-up (two patients), and technical limitations (one patient). Ten patients remain eligible for further therapy. The median survival of all assessable patients from the time of diagnosis of the primary tumor was 302 days (range, 0 to 70+ months) and the median survival from the institution of therapy was 207 days (range, 0 to 37+ months) (Fig 3). Presently, 14 patients remain alive at 3 to 37+ months after their first chemoembolization. Every patient had transient increases in liver function tests, particularly in lactic dehydrogenase. Almost every patient experienced some degree of nausea and/or vomiting, pain, and fever (Table 3). Peripheral edema developed in 14 patients and was accompanied by the development or exacerbation of ascites in 12 patients. In all but three patients, these symptoms resolved. Transient encephalopathy developed in three patients, and three had increases of creatinine greater than twofold, one of whom required temporary hemodialysis. No patients had significant hematologic toxicity. Six patients had significant complications (Table 4), two of which resulted in death; one exsanguinated from his tumor, and the other died from liver failure. DISCUSSION
The therapy of HCC has been extremely disappointing. Systemic chemotherapy produces a response rate of no more than 20% and does not impact favorably on the median survival of 3 to 6 months for the average patient with HCC. 4 Hepatic artery infusional chemotherapy may be more active, but has failed to improve survivor-
Downloaded from ascopubs.org by McMaster University on September 14, 2019 from 130.113.111.210 Copyright © 2019 American Society of Clinical Oncology. All rights reserved.
CHEMOEMBOLIZATION FOR HEPATOCELLULAR CARCINOMA
Fig 1.
1111
(A) Pretreatment CT scan and (B) CT scan 8 months after staged chemoembolization of a patient with HCC,
representing a partial response.
ships'6 and is associated with significant toxicity, particularly in patients with underlying liver disease. 29 Hepatic dearterialization accomplished by the ligation of the hepatic artery has been attempted in patients with HCC. No discernible improvement was seen in survival, and it has an operative mortality rate of as high as 23%.30 One-shot hepatic artery occlusion with particles alone has induced clinical responses in some patients, but the actual activity is difficult to measure due to the lack of rigorous response criteria in some trials.14,20 '22,31 Our pilot experience with 17 patients with a variety of hepatic neoplasms who were treated with particle embolization alone yielded little evidence of tumor necrosis, no change in liver function tests, and no pain following treatment. Although polyvinyl alcohol embolization without chemotherapy is active against
islet cell tumors, 16 it is our impression that the addition of chemotherapy was necessary in the treatment of HCC. The majority of the experience combining chemotherapy and particle embolization to treat hepatic tumors comes from Asia, where the high incidence of primary liver cancer has led to vigorous investigation of novel treatment modalities. In our trial, Gelfoam powder (diameter of 40 to 50 gm) was used as the embolizing material with the expectation that it would produce a more distal arteriolar/capillary occlusion with less collateralization to the tumor than would occur with larger Gelfoam particles or unprocessed polyvinyl alcohol (250 to 590 am). Gelfoam undergoes degradation and resorption in the circulation, and the vessels recanalize within 48 to 72 hours,32 while undegradable polyvinyl alcohol causes permanent vascular occlusion. 33
Fig 2. (A) Pretreatment CT scan and (B)CT scan 8 months after whole-liver chemoembolization for HCC. This demonstrates the appearance of liquefaction necrosis without evidence of tumor shrinkage.
Downloaded from ascopubs.org by McMaster University on September 14, 2019 from 130.113.111.210 Copyright © 2019 American Society of Clinical Oncology. All rights reserved.
1112
VENOOK ET AL Table 4. Complications of Chemoembolization (N = 51)
1
Type of Complication
No. of Patients
Tumor hemorrhage
1
Tumor rupture
1
Liver failure Cerebral hemorrhage Liver abscess Tumor lysis syndrome
1 1 1 1
(j~ 4-,
0.
U
Time (months) Fig 3. Actuaria I survival of all assessable patients with HCC from the time of initiation of chemoembolization.
This difference allows for repeated treatments of the same vascu lar bed with Gelfoam, whereas with polyvinyl a alcohol, flow is permanently redistributed into col lateral vessels. The most ac tive chemotherapeutic agents in HCC are fluoro uracil, doxorubicin, and mitomycin. Meaningful doses of fluorouracil could not be incorporated into the chemoembolization mixture due to the low drug concentration of the available soluti )n. Cisplatin was combined with doxorubicin and mitomycin based on the unpublished pilot expterience of Aigner et al (personal communication. February 1986), who found the treatment to be active and well-tolerated in a few Table 3. Toxic ity of Chemoembolization (N Type of To'Xicity xic, Transient LFT ele vation Pain Nausea and/or vomiting Fever Edema Ascites Transient encephhalopathy Alopecia Transient renal d ysfunction* Infection Hematologic
=
51)
patients. Since cisplatin, mitomycin, and doxorubicin have relatively slow total body clearances and limited hepatic extractions, they are less than ideal for unmodified regional infusion. 34 However, enhanced regional drug exposure may be achieved when IA administration of these drugs is combined with the interruption of arterial blood flow, as has been demonstrated with chemoembolization of coloreccisplatin/collagen tal metastases 27 and with mitomycin when admin35 istered with degradable starch microspheres. While the radiographically documented PR rate observed in this trial is only 24%, the
appearance of liquefaction on CT scan, as occurred in 70% of our patients, has been shown to correlate histopathologically with significant tumor necrosis.'5, 24 This has been confirmed in four of our patients who went to surgery following chemoembolization and were found to have mostly necrotic tumor. In other words, liquefaction necrosis may represent a clinically significant antitumor effect (Figs 1 and 2). Therefore, applying standard radiographic response criteria to judge this type of treatment may underestimate its activity. We feel that this treatment had a discernible and favorable impact on our patients. Since we serve as a referral center, our patients tend to have advanced disease (a median of 50% liver replacement). Furthermore, a majority of our
"LI 'r"'"
T
"rr c"rr "r'-' 'Y lllll·P 51nlllrPn No. of (%positive and/or had evidence of cirrhosis, charac51 (100) teristics that portend a bad outcome regardless of 48 (94) the treatment rendered. 29 The median survival 47 (92) for patients with HCC and cirrhosis may be as 46 (90) 14(27) little as 1.2 months.' Doci et a129 reported a jJIAIU
V
i T.' Lfh
f
I
i~ L LW
AS04
11
1L1
No. of Patients Patients
12(24)
median survival of 3.5 months in a group of
3 (6) 3 (6)
patients similar to ours who were treated with IA
3(6)
chemotherapy. Therefore, a median survival of 7
(2) 0(0)
months from first treatment, with 25% of the patients surviving more than 20 months from the time of treatment, seems significant. The largest experience with chemoemboliza-
1
Abbreviation: LFT, liver function test. *Twofold or greatetr increase in serum creatinine.
Downloaded from ascopubs.org by McMaster University on September 14, 2019 from 130.113.111.210 Copyright © 2019 American Society of Clinical Oncology. All rights reserved.
CHEMOEMBOLIZATION FOR HEPATOCELLULAR CARCINOMA
tion for HCC comes from Asia.15,18-2 4 However, differences in patient selection and methodology, including choice of drugs and particle, the route of drug delivery, the inclusion of Lipiodol UltraFluide (Guerbet, Aulnay sous Bois, France) in the mixture, 23 and the sequence in which these are administered, make it difficult to compare those results with ours. In Japan, chemoembolization is also used preoperatively in patients with resectable tumors in an attempt to decrease tumor vascularity and the risk of hematogenous metastases at the time of liver mobilization. No benefit for this approach has been shown.36 Toxicity was significant but predictable and mostly self-limited. In general, patients who received whole liver chemoembolization had more pronounced toxicity than those treated with selective procedures. Although virtually every patient experienced nausea, vomiting, pain, and fever, the majority were ambulatory within 4 days and largely asymptomatic within 2 weeks. Patients' quality of life was further aided by the absence of repeated chemotherapy treatments until tumor progression became evident. The median hospital stay of 4 days is less than that for a single standard cycle of IA therapy. Although Takayasu et a137 reported a 50% incidence of postembolization cholecystitis due to occlusion of the cystic artery, we encountered no clinically apparent gall bladder complications. Since we used no
1113
special precautions to avoid embolizing the cystic artery, this discrepancy with Takayasu et al's experience is unexplained. The two treatmentrelated deaths occurred in patients with very little hepatic reserve (one with extensive cirrhosis and one with a large, necrotic tumor). Our experience in this trial leads us to conclude that patients with HCC who are good candidates for chemoembolization are those who have unresectable, liver-predominant disease, a Karnofsky performance status of 60% or better, no ascites, and adequate hepatic reserve (albumin > 2.5 g/dL, bilirubin < 3 mg/dL, normal prothrombin time, and partial thromboplastin time). While this trial demonstrates encouraging activity of chemoembolization in HCC, the following questions regarding this modality remain unresolved: (1) Should re-treatment be used at set intervals rather than at the time of hepatic progression, as used in this trial? (2) Which of these or other drugs should be used in the chemoembolization mixture? (3) Is the selection of the specific embolization material important for a given tumor type? (4) How does one reproducibly and systematically describe the effects of this treatment given the experience that changes in tumor volume may not tell the entire story? While these questions are being addressed, we believe the treatment should remain confined to protocols at dedicated treatment centers.
REFERENCES 1. Smalley SR, Moertel CG, Hilton JF, et al: Hepatoma in the noncirrhotic liver. Cancer 62:1414-1424, 1988 2. Malt RA: Surgery for hepatic neoplasms. N Engl J Med 313:1591-1596, 1985 3. Starzl TE, Iwatsuki S, Shaw BW, et al: Treatment of fibrolamellar hepatoma with partial hepatectomy and transplantation of the liver. Surg Gynecol Obstet 162:145-148, 1986 4. Lewis BJ, Friedman MA: Current status of chemotherapy for hepatoma, in Ogawa M (ed): Chemotherapy of Hepatic Tumors. Princeton, NJ Excerpta Medica, 1984, pp 63-74 5. Wellwood JM, Cady B, Oberfield RA: Treatment of primary liver cancer: Response with regional chemotherapy. J Clin Oncol 5:25-31, 1979 6. Urist MM, Balch CM: Intraarterial chemotherapy for hepatoma using Adriamycin administered via the implantable infusion pump. Proc Am Soc Clin Oncol 3:146, 1983 (abstr) 7. Order SE, Stillwagon GB, Klein JL, et al: Iodine 131 antiferritin, a new treatment modality in hepatoma: A Radiation Therapy Oncology Group study. J Clin Oncol 3:1573-1582, 1985
8. Order S, Pajak T, Klein J, et al: A randomized prospective trial in nonresectable hepatoma comparing Adriamycin and 5-fluorouracil + 131-I antiferritin: An RTOG study. Proc Am Soc Clin Oncol 8:99, 1989 (abstr) 9. Breedis C, Young G: The blood supply of neoplasms in the liver. Am J Pathol 30:969-985, 1954 10. Markowitz J: The hepatic artery. Surg Gynecol Obstet 95:644-646, 1952 11. Sparks F, Mosher M, Hallauer W, et al: Hepatic artery ligation and postoperative chemotherapy for hepatic metastases: Clinical and pathophysiological results. Cancer 35:1074-1082, 1975 12. Charnsangavej C, Chuang VP, Wallace S, et al: Angiographic classification of hepatic arterial collaterals. Radiology 144:485-494, 1982 13. Goldstein HM, Wallace S, Anderson JH, et al: Transcatheter occlusion of abdominal tumors. Radiology 120:539-545, 1976 14. Clouse ME, Lee RGL, Duszlak EJ: Peripheral hepatic artery embolization for primary and secondary hepatic neoplasms. Radiology 147:407-411, 1983 15. Yamada R, Sato M, Kawabata M, et al: Hepatic
Downloaded from ascopubs.org by McMaster University on September 14, 2019 from 130.113.111.210 Copyright © 2019 American Society of Clinical Oncology. All rights reserved.
1114
VENOOK ET AL
artery embolization in 120 patients with unresectable hepatoma. Radiology 148:397-401, 1983 16. Ajani JA, Carrasco H, Charnsangavej C, et al: Islet cell tumors metastatic to the liver: Effective palliation by sequential hepatic artery embolization. Ann Intern Med 108:340-344, 1988 17. Fujimoto S, Miyazaki M, Endoh F, et al: Biodegradable mitomycin c microspheres given intra-arterial for inoperable hepatic cancer. Cancer 56:2404-2410, 1985 18. Ohnishi K, Tsuchiya S, Nakayama T, et al: Arterial chemoembolization of hepatocellular carcinoma with mitomycin c microcapsules. Radiology 152:51-55, 1984 19. Sakurai M, Okamura J, Kuroda C: Transcatheter chemoembolization effective for treating hepatocellular carcinoma. Cancer 54:387-392, 1984 20. Lin D, Liaw Y, Lee T, et al: Hepatic arterial embolization in patients with unresectable hepatocellular carcinoma-A randomized controlled trial. Gastroenterology 94:453-456, 1988 21. Soga K, Nomoto M, Ichida T, et al: Clinical evaluation of transcatheter arterial embolization and one-shot chemotherapy in hepatocellular carcinoma. Hepatogastroenterology 35:116-120, 1988 22. Sato Y, Fujiwara K, Ogata I, et al: Transcatheter arterial embolization for hepatocellular carcinoma. Cancer 55:2822-2825, 1985 23. Shimamura Y, Gunven P, Takenaka Y, et al: Combined peripheral and central chemoembolization of liver tumors. Cancer 61:238-242, 1988 24. Takayasu K, Moriyama N, Muramatsu Y, et al: Hepatic arterial embolization for hepatocellular carcinoma. Radiology 150:661-665, 1984 25. Carrasco C, Wallace S, Charnsangavej C, et al: Treatment of hepatic metastases in ocular melanoma. JAMA 255:3152-3154, 1986 26. Daniels J, Daniels A, Quinn M, et al: Phase I trial with cisplatin or mitomycin hepatic chemoembolization with Angiostat collagen for embolization in patients with colorectal cancer. Proc Am Soc Clin Oncol 7:385, 1988 (abstr)
27. Modiano M, Holdsworth M, Alberts D, et al: Control of hepatic metastases from colorectal carcinoma with selective intraarterial chemoembolization with Angiostat collagen, CDDP, mitomycin and Adriamycin. The University of Arizona experience. Proc Am Soc Clin Oncol 8:105, 1989 (abstr) 28. Venook A, Chase J, Stagg R, et al: A phase I-II trial of angiostat/mitomycin-c hepatic chemoembolization for colorectal cancer metastatic to the liver. Proc Am Soc Clin Oncol 8:120, 1989 (abstr) 29. Doci R, Bignami P, Bozzetti F, et al: Intrahepatic chemotherapy for unresectable hepatocellular carcinoma. Cancer 61:1983-1987, 1988 30. Almersjo O, Bengmark S, Rudenstam CM, et al: Evaluation of hepatic dearterialization in primary and secondary cancer of the liver. Am J Surg 124:5-9, 1972 31. Allison DJ, Jordan H, Hennessy O: Therapeutic embolisation of the hepatic artery: A review of 75 procedures. Lancet 1:595-599, 1985 32. Barth KH, Strandberg JD, White RI: Long term follow-up of transcatheter embolization with autologous clot, oxycel and Gelfoam in domestic swine. Invest Radiol 12:273280, 1977 33. Chuang VP, Wallace S, Soo CS, et al: Therapeutic ivalon embolization of hepatic tumors. Am J Roent 138:289294, 1982 34. Collins JM: Pharmacologic rationale for regional drug delivery. J Clin Oncol 2:498-504, 1984 35. Andersson M, Aronsen KF, Balch C, et al: Pharmacokinetics of intra-arterial mitomycin C with or without degradable starch microspheres (DSM) in the treatment of nonresectable liver cancer. Acta Oncol 28:219-222, 1989 36. Nagasue N, Galizia G, Kohno H, et al: Adverse effects of preoperative hepatic artery chemoembolization for resectable hepatocellular carcinoma: A retrospective comparison of 138 liver resections. Surgery 106:81-86, 1989 37. Takayasu K, Moriyama N, Muramatsu Y, et al: Gallbladder infarction after hepatic artery embolization. AJR 144:135-138, 1985
Downloaded from ascopubs.org by McMaster University on September 14, 2019 from 130.113.111.210 Copyright © 2019 American Society of Clinical Oncology. All rights reserved.
