CA 125: a clinically useful tumor marker in the management of colorectal carcinoma metastatic to the liver in patients with normal carcinoembryonic antigen.

Two patients with colon carcinoma metastatic to the liver had normal plasma carcinoembryonic antigen (CEA) levels (<1.0 ng/ml) but elevated CA 125 levels. Treatment of the metastatic disease with chemotherapy, plus surgery in one case, led to declines in the CA 125 levels. These decreases were associated with tumor regression, as confirmed by clinical and radiologic evidence. These findings lead us to conclude that the measurement of CA 125 for patients with normal CEA levels is useful in the management of colorectal carcinoma.

Hepatic arterial infusion of recombinant platelet factor-4 suppresses metastases to the lungs from tumors implanted into the livers of rabbits.

BACKGROUND: This study evaluated the toxicity (Part I) and antitumor effects (Part II) associated with hepatic arterial infusion of recombinant platelet factor-4 (rPF4), an antiangiogenic protein. METHODS: Healthy rabbits (Part I) and rabbits with tumors implanted in their livers (Part II) received saline or rPF4 via hepatic arterial infusion. Three saline-receiving and four rPF4-receiving animals died 2-3 days postinfusion from gastroduodenal thromboembolism. The remaining animals were necropsied 3, 7, 10, or 14 days postinfusion. Blood analyses and hepatic angiography were performed before infusion and at the time of sacrifice. RESULTS: In Part I, focal coagulation necrosis of the hepatic parenchyma was observed in 1 of 11 rabbits that received saline and in 6 of 10 that received rPF4. In Part II, hepatic arterial infusion of rPF4 had no effect on growth of the implanted liver tumors. However, the protein significantly reduced the incidence of lung metastasis. CONCLUSIONS: Intraarterial infusion of rPF4 significantly reduced the incidence of lung metastasis. Nonheparin systemic anticoagulation may be needed during catheterization and infusion procedures to prevent thromboemboli.

Gastrointestinal leiomyosarcoma metastatic to the liver. Durable tumor regression by hepatic chemoembolization infusion with cisplatin and vinblastine.

BACKGROUND: Gastrointestinal leiomyosarcoma metastatic to the liver is considered most resistant to any combination of systemic chemotherapy containing doxorubicin and/or ifosphamide. METHODS: Fourteen patients with gastrointestinal leiomyosarcoma metastatic to the liver were treated with hepatic chemoembolization infusion consisting of polyvinyl alcohol sponge particles mixed with cisplatin powder (150 mg) followed by an intrahepatic arterial infusion of vinblastine (10 mg/m2). RESULTS: Ten major (> 50% regression) tumor responses were observed (70%) in patients lasting from 8 to 31+ months (median, 12 months) after an average of two hepatic chemoembolization procedures, usually 4 weeks apart. Transient side effects included right upper quadrant pain requiring narcotics, significant hepatic enzyme elevation, particularly of lactic dehydrogenase with a minimal increase in bilirubin, paralytic ileus requiring nasogastric suction up to 72 hours, urinary electrolyte losses (potassium+, magnesium++, sodium+) requiring supplements, and occasionally mild but transient leukopenia and thrombocytopenia. CONCLUSIONS: Hepatic chemoembolization infusion appears to induce a high rate of durable tumor response in patients with notoriously chemoresistant gastrointestinal leiomyosarcoma metastatic to the liver.

Durable hepatic tumor regression after arterial chemoembolization-infusion in patients with islet cell carcinoma of the pancreas metastatic to the liver.

BACKGROUND: Islet cell carcinoma of the pancreas is a neuroendocrine tumor often presenting with left upper quadrant mass and radiographic evidence of liver metastases. Because survival among these patients is determined largely by the pace of metastatic events in the liver, significant palliation may be achieved by regional hepatic therapy. METHODS: Five patients with islet cell carcinoma of the pancreas metastatic to the liver (four nonfunctional, one gastrin producing), were treated by hepatic arterial chemoembolization-infusion consisting of a mixture of polyvinyl alcohol sponge (150 mg) and cisplatin (150 mg) followed by 2-hour intraarterial infusion of vinblastine (10 mg/m2). Each patient received two such treatments, 1 month apart, requiring 3 to 6 days of hospital admission. RESULTS: Significant tumor regression (> 50%) was observed in four of five patients, lasting from 8 to 44 months. Toxicity was limited to right upper quadrant pain, paralytic ileus requiring nasogastric suction for 24 to 72 hours, transient, mild bilirubinemia and liver enzyme elevation, hypomagnesemia and hypokalemia, and occasionally, moderate, self-limiting granulocytopenia. CONCLUSIONS: This preliminary, albeit limited, experience with hepatic chemoembolization-infusion in patients with islet cell carcinoma metastatic to the liver emphasizes the high incidences of durable tumor regression that can be achieved with minimal iatrogenic intervention.

Regional biologic therapy. Hepatic arterial infusion of recombinant human tumor necrosis factor in patients with liver metastases.

Twenty-two chemotherapy-resistant patients with liver metastases received 46 courses of recombinant human tumor necrosis factor (rhTNF) administered by 5-day continuous infusion through percutaneously inserted hepatic arterial catheters. The maximum tolerated daily dose of rhTNF was 150 micrograms/m2. This is six times the maximum tolerated daily dose of rhTNF that could be given systemically (intravenous) on the same schedule. The dose-limiting toxicity resulted in severe, although transient, hypophosphatemia (less than 1.0 mg/dl) associated with myocardial dysfunction. Objective tumor response (partial tumor response or greater) was observed in 2 of 14 patients (14%) with colorectal cancer and lasted as long as 3 months. Three additional minor responses occurred among these patients with colorectal cancer. Plasma carcinoembryonic antigen levels also decreased significantly (greater than 25%) in 7 of the 14 (50%) patients with colorectal cancer. Regional biologic therapy with rhTNF as a sole modality has definite antitumor activity in colorectal cancer metastatic to the liver and warrants additional study in previously untreated patients.

The prevalence and location of metastases from ocular melanoma: imaging study in 110 patients.

Ocular melanoma is characterized by an unpredictable clinical course, during which fulminant metastatic disease may occur after a prolonged disease-free interval. The purpose of this study was to determine the pattern of metastatic involvement in this disease. The clinical and radiologic findings in 110 patients with metastatic ocular melanoma were reviewed. The 54 men and 56 women were 24-79 years old (mean, 50 years) when the primary tumor was first diagnosed. Metastases were present in three patients at the time of first diagnosis and occurred in 107 patients 2 months to 36 years later (mean, 52 months). One hundred five patients died between 1 and 38 months after the onset of metastatic disease. Hepatic metastases developed in 101 patients (92%), and in 60 (55%) of these, the liver was the only organ involved initially. Pulmonary parenchymal metastases developed in 34 patients (31%), but in only four of them were metastases confined to the lungs. Twenty-five patients (23%) had bone involvement, mostly affecting the spine. Nineteen patients (17%) had skin or subcutaneous metastases, but in only two of them was this the initial finding. Nodal involvement was shown in 15 patients (14%), almost always associated with extensive hepatic metastases. Brain and adrenal metastases were seen in five and three patients, respectively. Hepatic involvement occurs in almost all patients who develop metastatic ocular melanoma, and the liver is the most common initial site of metastatic involvement. Metastases may develop after a long disease-free interval.

Arterial chemotherapy in the management of colorectal cancer: an overview.

We have discussed the role of arterial therapy in patients with various stages and types of colon cancer. Arterial therapy is probably not useful as an adjuvant therapy for Dukes' C colon cancer. It may, however, play a role among patients with incomplete resection of liver metastases (positive margins). A randomized trial is needed to determine the role of arterial therapy in patients who have undergone complete resection of liver metastasis. Arterial therapy does not seem justified for patients with recurrent pelvic tumors. For nonresectable liver metastases, hepatic arterial therapy induces a higher response rate than does intravenous treatment. It may also improve performance status and offer additional palliation to patients who have failed systemic chemotherapy. are refractory to systemic chemotherapy may be candidates for palliative hepatic arterial chemotherapy even out of the context of a clinical trial. Asymptomatic patients with nonresectable liver metastasis who are refractory to systemic chemotherapy should be enrolled in phase I-II arterial chemotherapy trials designed to identify optimal treatment regimens. Previously untreated asymptomatic patients wishing treatment may be enrolled in a new multi-institutional phase III trial being designed to compare contemporary systemic chemotherapy with less toxic arterial therapy and combined arterial and systemic therapy. Such a new trial will have to avoid any cross-over between arms to determine the true impact of arterial therapy on survival. Regional arterial chemotherapy tries to extract the "extra mile" from marginally active drugs that have a steep dose response curve by increasing tumor drug exposure. Increased drug concentrations in the tumor may be accomplished by means of the blood vessel-to-tumor concentration gradient. The technology to achieve such a gradient has involved percutaneous hepatic arterial catheters, implantable infusion pumps or ports, and external pumps. The most economic hepatic arterial delivery system for protracted arterial FUdR is an infusion pump. Despite good pharmacological rationale, an improved response rate, and good evidence for effective palliation in advanced disease, hepatic arterial therapy has not improved survival when compared with systematic intravenous treatment. Possible explanations include the following: (1) poor study design that allowed patients to cross over between arms: (2) inadequate arterial chemotherapy combination; (3) inadequate arterial chemotherapy schedule; (4) hepatobiliary toxicity levels that required cessation of hepatic arterial therapy and allowed the emergence of resistant tumor clones; and (5) systemic progression of disease. Only time will tell whether improved chemotherapy and the design of a new phase III trial will establish a beneficial role for upfront hepatic arterial therapy in asymptomatic patients with colon cancer metastatic to the liver.

Regression of hepatic metastases from gastrointestinal leiomyosarcoma after hepatic arterial chemoembolization.

Two patients with gastrointestinal leiomyosarcoma metastatic to the liver were treated by hepatic chemoembolization with cisplatin and polyvinyl sponge followed by hepatic arterial infusion of vinblastine. Effective palliation in terms of durable tumor regression was achieved in both patients after two chemoembolization-infusion procedures. These results suggest that regional therapy may offer new hope for the subset of sarcoma patients who have liver metastases resistant to combination systemic chemotherapy.

Severe, symptomatic, dose-limiting hypophosphatemia induced by hepatic arterial infusion of recombinant tumor necrosis factor in patients with liver metastases.

Twenty-two patients with liver metastases received 45 courses of recombinant tumor necrosis factor (rTNF) by hepatic arterial infusion in doses ranging from 12.5 to 175 micrograms/m2/d for 5 days by continuous infusion. The induction of statistically significant, dose-related, severe, albeit transient, hypophosphatemia (less than 1.0 mg/dl) associated with clinically significant, right-sided myocardial dysfunction and severe lassitude was observed. These side effects were promptly reversed after rTNF was stopped and intravenous phosphate supplementation was started. As no significant or consistent increase in urinary phosphate excretion was detected, the rTNF-induced hypophosphatemia probably resulted from an intracellular shift of phosphate. Since tumor regression was clearly associated with the lowest levels of serum phosphate, hypophosphatemia may be important in the antitumor effects of rTNF.

Increased incidence of hypersensitivity to iodine-containing radiographic contrast media after interleukin-2 administration.

Eight of 28 (28%) cancer patients with liver metastases treated by either splenic (four) or hepatic (four) arterial infusion of recombinant interleukin-2 (rIL-2) developed hypersensitivity reactions to iodine-containing radiographic contrast media. These reactions consisted of fever, chills, malaise, nausea and vomiting, skin rash, diarrhea, and occasionally, hypotension. Reactions usually occurred 1 month after the initial arteriographic procedure and rIL-2 infusion, with 1-hour to 4-hour intervals between procedure and reexposure of the patient to the iodine-containing contrast medium (used in conjunction with computerized tomography or repeated arteriography for subsequent courses of rIL-2 infusions) and the onset of symptoms. Prompt administration of corticosteroids during the reaction and premedication of patients who were known to have had a reaction in the past were very effective in stopping reactions or preventing them from reoccurring. The high incidence (28%) of hypersensitivity reactions, the temporal relationship (4 hours) between the arteriographic procedure (utilizing iodine-containing contrast medium) and the initial infusion of rIL-2 (while some of the contrast medium was still present), and the absence of such hypersensitivity reactions among patients receiving systemic (intravenous) rIL-2 (not requiring the use of concomitant iodine-containing contrast medium) provide additional evidence that in the presence of a potentially immunogenic moiety, rIL-2, a potent stimulant of the human immune system, can produce an initial sensitization followed by subsequent anamnestic reaction upon reexposure of the patient to the immunogen (even without the additional rIL-2).

Splenic versus hepatic artery infusion of interleukin-2 in patients with liver metastases.

In an attempt to improve the therapeutic index of recombinant interleukin-2 (rIL-2) by generating or activating lymphokine-activated killer (LAK) cells and tumor-infiltrating lymphocytes (TIL) regionally and/or in situ, we randomly assigned 28 patients with liver metastases to receive rIL-2 by continuous infusion for 5 days via either the splenic artery or the hepatic artery. Clinically significant and lasting tumor regression was observed only in two of 28 patients (7%), one in each of the two treatment arms. The maximum-tolerated daily dosage of rIL-2 was 3 x 10(6) U/m2; beyond this dosage, toxicity was excessive. Peripheral LAK cell activity measured in vitro and clinical tumor regression did not correlate. This observation, coupled with the equal distribution of regressions between the two treatment arms, raises the possibility that tumor regression, rare though it may be in response to rIL-2 administration, is largely mediated by TIL activation and not by LAK cell generation.

Hepatic arterial infusion of human recombinant tumor necrosis factor-alpha. An experimental study in dogs.

Local and systemic toxicities associated with hepatic arterial infusion of human recombinant tumor necrosis factor (rTNF) were studied in healthy adult mongrel dogs. The animals received saline containing human serum albumin with or without rTNF (0.02, 0.2, or 2.0 mg/m2). Arteriograms were made, and blood samples were collected for hematologic and biochemical analyses at regular intervals. The dogs were killed at 1, 3, and 7 days postinfusion and complete necropsies were performed. Specimens were obtained from various tissues for histopathologic evaluation. Results indicated that all but the highest dose of rTNF were well tolerated. Severe histopathologic changes were found in the liver, spleen, and kidneys of the animals receiving 2.0 mg/m2 rTNF. In addition, focal tubular degeneration was found in one dog administered 0.2 mg/m2 rTNF. These data suggest that the upper dose limit for hepatic arterial infusion of rTNF is between 0.2 and 2.0 mg/m2 and that renal function should be closely monitored after infusion.

Regression of ocular melanoma metastatic to the liver after hepatic arterial chemoembolization with cisplatin and polyvinyl sponge.

Thirty patients with ocular melanoma metastatic to the liver were treated by hepatic arterial chemoembolization using an admixture of cisplatin and polyvinyl sponge. Tumor regression was complete in one patient and partial (greater than 50%) in 13 patients. The total response rate was 46%. The median survival for the entire group was 11 months (95% confidence interval, nine to 18 months). Treatment-related morbidity was short-lived and included primarily severe upper right quadrant abdominal pain, transient paralytic ileus, and nonicteric hepatitis. Hepatic arterial chemoembolization provided effective palliation, with good-quality survival among 46% of patients with ocular melanoma metastatic to the liver.

Fractionated extract of Astragalus membranaceus, a Chinese medicinal herb, potentiates LAK cell cytotoxicity generated by a low dose of recombinant interleukin-2.

Success with rIL-2 immunotherapy of human cancer appears to depend on the administration of high doses which are frequently associated with excessive toxicity. Future use of rIL-2 will require certain modifications based on the use of lower doses of rIL-2 without significant loss of antitumor efficacy. We tested in vitro the possibility of potentiating the activity of rIL-2 in terms of LAK cell generation. We hypothesized that co-incubation of LAK cell precursors with a Chinese herbal extract (F3) of Astragalus membranaceus, (an immune modulator currently under study in our laboratory), along with a low concentration of rIL-2, would generate levels of LAK cell activity equivalent to those generated by high concentrations of rIL-2 alone. We found (1) a 10-fold potentiation of rIL-2 activity manifested by tumor cell-killing activity of 80% resulting from LAK cell generation with F3 plus 100 u/ml of rIL-2 versus 76% generated by 1,000 u/ml of rIL-2 alone; (2) a significant reduction in the number of effector LAK cells required for equicytotoxic reaction following LAK cell generation with F3 plus rIL-2 compared to rIL-2 alone. We conclude that potentiation of antitumor activity mediated by rIL-2 in low concentrations is possible by the concomitant use of another immune modulator such as Astragalus membranaceus.

Hepatocellular carcinoma. A retrospective analysis of treatments to manage disease confined to the liver.

The histories of 73 patients with hepatocellular carcinoma (HCC) confined to the liver who were seen at M. D. Anderson Hospital between January 1976 and December 1983 were reviewed. In 18 patients the tumor was resected either at the outset or after the patients' response to chemotherapy; nonsurgical treatments consisted of hepatic arterial infusion (HAI) in 10 patients and intravenous (IV) therapy in four patients. Patients who had resection were younger, and their liver functions and performance status were better than the IV and HAI groups. Their median survival was 46 months. Of the patients who had nonresectable tumors, 28 received chemotherapy by HAI and 27 received IV therapy. Of the 28 patients in the HAI treatment group, 25 received uniform infusion of floxuridine (FUDR, Roche, Australia), doxorubicin (Adriamycin), and mitomycin C (FUDRAM). Of the 27 patients in the IV treatment group, 15 received 5-fluorouracil (5-FU) and doxorubicin-containing regimens; in 11 patients 5-FU was combined with other agents. The HAI and IV treatment groups were similar in age and ethnicity, performance status, serum alpha-fetoprotein levels, liver function, presence of hepatitis B antigen, and presence or absence of cirrhosis. The median survival was 9 months for HAI-treated patients and 5 months for the IV-treated group. The statistical differences were resection versus HAI, P less than 0.01; resection versus IV, P less than 0.01; HAI versus IV, P less than 0.01. Thirteen of 18 patients who had resections, six of 28 patients treated with HAI, and two of 27 IV-treated patients survived 2 years or more. It is concluded that for patients with hepatocellular carcinoma confined to the liver, the option of tumor resection either at the beginning of treatment or after chemotherapy offers the best chances for long-term survival. The overall prognosis is poor for patients with nonresectable hepatocellular carcinoma, but arterial infusion chemotherapy may double the median survival as compared to IV chemotherapy.

Immunotherapy with Chinese medicinal herbs. II. Reversal of cyclophosphamide-induced immune suppression by administration of fractionated Astragalus membranaceus in vivo.

A partially purified fraction (F3) with an estimated molecular weight of 20,000 to 25,000 derived from the traditional Chinese medicinal herb Astragalus membranaceus, was found to possess a potent immunorestorative activity in vitro. Its capacity to aborogate the local xenogeneic graft versus host reaction (XGVHR) following injection in vivo was further studied in a newly developed animal model designed for preclinical evaluation of various biological response modifiers. F3 was injected intravenously into cyclophosphamide-primed rats at varied concentrations and schedules prior to grafting of mononuclear cells from healthy normal donors. Maximal abrogation of the local XGVHR mounted by the mononuclear cells, was observed following injection of 5.55 mg of F3 daily for eight days. This abrogation of XGVHR indicates a reversal of the immunosuppressive effect of cyclophosphamide as manifested by a significant decline in the local XGVHR volume from 99.42 +/- 9.2 mm3 (positive control) to 39.78 +/- 8.3 mm3 (p less than 0.001). This reversal of cyclophosphamide-induced immunosuppression by the administration of F3 was complete, since the volume of the abrogated local XGVHR (39.78 +/- 8.3 mm3) was comparable to 34.79 +/- 5.69 mm3 (p greater than 0.1) in the negative control group (no cyclophosphamide-priming; saline injection only). These data indicate that F3 administration markedly enhances the rats' ability to reject the xenogeneic graft and therefore possesses a strong immune potentiating activity in vivo. These preclinical data also provide the rational basis for the use of extracts of Astragalus membranaceus in phase I clinical trials among patients suffering from iatrogenic or inherent immune deficiency states.

Immunotherapy with Chinese medicinal herbs. I. Immune restoration of local xenogeneic graft-versus-host reaction in cancer patients by fractionated Astragalus membranaceus in vitro.

The in vitro immunomodulatory activity of fractions derived from Astragalus membranaceus, an herb commonly used in the practice of traditional Chinese medicine, was first screened by studying their individual effects on mononuclear cells (MNC) derived from healthy normal donors using the local xenogeneic graft-versus-host reaction (XGVHR). Sephacryl S-200 column-separated Fraction 3 (MW 20,000-25,000) along with its crude extract precursor, Fraction 7, and another crude extract derivative, Fraction 8, were equally augmentative (p less than 0.05) in their effect on MNC from normal donors. These three active fractions were further studied on MNC derived from 13 cancer patients. Using again the local XGVHR as a model assay for T-cell function, preincubation of MNC derived from cancer patients with Fraction 3 induced a significant increase in local XGVHR (compared to untreated cells) with a mean +/- SD of 151.34 +/- 46.02 mm3 vs 57.80 +/- 16.44 mm3; p less than 0.001. Fractions 7 and 8 likewise induced significant increases in local XGVHR (109.14 +/- 19.32 mm3 versus 50.91 +/- 17.39 mm3; p less than 0.001 and 119.74 +/- 18.33 mm3 versus 48.77 +/- 16.17 mm3; p less than 0.001, respectively). The augmented immune reactions which were induced by either Fraction 3 or Fraction 8 (but not by Fraction 7) in MNC derived from cancer patients, each significantly exceeded the local XGVHR observed in the untreated MNC derived from normal donor controls with a relative reference index (ratio) of 1.60 +/- 0.48 and 1.23 +/- 0.17 respectively; p less than 0.005.(ABSTRACT TRUNCATED AT 250 WORDS)

Abrogation of the local xenogeneic graft versus host reaction by preinjection of cimetidine to the immunosuppressed host rats: an experimental model for testing preclinical immunorestorative activity in vivo.

The local xenogeneic graft-vs-host reaction (XGVHR) is mediated by competent T lymphocytes. It was effectively used as a practical bioassay to assess the in vitro immunomodulatory effects of several drugs on cells derived from cancer patients. A modified version of the XGVHR model was recently developed primarily for preclinical evaluation of in vivo immunorestoration induced by experimental biological response modifiers (BRMs). A well-defined biological response modifier, cimetidine, was injected into Cytoxan-primed rats prior to their inoculation with xenogeneic human mononuclear cells. The cimetidine treatment induced a partial abrogation of the immunosuppressive effect of Cytoxan as manifested by a significant decline in the volume of the XGVHR from 115.23 +/- 15.72 mm3 (positive control) to 67.3 +/- 11.41 mm3 (P less than 0.01). This abrogation of Cytoxan-induced immunosuppression by cimetidine was incomplete since the XGVHR without Cytoxan (negative control, saline only) was still significantly lower (45.12 +/- 4.55 mm3; P less than 0.01). The effect of cimetidine injection in vivo appeared to be dose dependent and did not exhibit any nonspecific toxic effect to the mononuclear cell inoculum. These results indicate that the model can serve as a useful tool in the preclinical evaluation of newly developed immunorestorative biological response modifiers.