Cooperative updating in the Hopfield model.

We propose a new method for updating units in the Hopfield model. With this method two or more units change at the same time, so as to become the lowest energy state among all possible states. Since this updating algorithm is based on the detailed balance equation, convergence to the Boltzmann distribution is guaranteed. If our algorithm is applied to finding the minimum energy in constraint satisfaction and combinatorial optimization problems, then there is a faster convergence than those with the usual algorithm in the neural network. This is shown by experiments with the travelling salesman problem, the four-color problem, the N-queen problem, and the graph bi-partitioning problem. In constraint satisfaction problems, for which earlier neural networks are effective in some cases, our updating scheme works fine. Even though we still encounter the problem of ending up in local minima, our updating scheme has a great advantage compared with the usual updating scheme used in combinatorial optimization problems. Also, we discuss parallel computing using our updating algorithm.

[Significance of anti-HBc IgM in acute hepatitis and hepatitis B associated chronic liver disease].

To determine the diagnostic value of anti-HBc IgM in acute viral hepatitis or chronic liver disease B, Anti-HBc IgM was measured by a RIA and an ELISA in 32 patients with acute hepatitis (4 with type A, 15 with type B and 13 with type non A non B), 18 patients with chronic hepatitis and 19 patients with liver cirrhosis. In acute hepatitis B, anti-HBc IgM (both RIA and ELISA) was positive in 14(93%) of 15 patients and its cut-off index value was very high. However, anti-HBc IgM was always negative in one patient with typical course of type B. In 1 of 4 patients with acute hepatitis A and 2 of 13 with non A non B, anti-HBc IgM (RIA and/or ELISA) was positive. These 3 patients were positive for anti-HBs at the onset of disease, so we could not made the diagnosis of acute hepatitis B. Anti-HBc IgM was positive in 21(51%) of 37 patients with HBsAg-positive chronic liver disease by RIA and in 11 (30%) by ELISA, and its cut-off index value was relatively low. These results suggest that when adequate cut-off index value is established, anti-HBc analysis is useful for differentiating recent and current infections from remote infections.

Significance of HB virus infection in an area of Japan with high incidence of liver cirrhosis and hepatocellular carcinoma: an analysis of consecutive studies among inhabitants of Tomié-Town, Goto Islands.

In an attempt to clarify the relationship of liver cirrhosis and hepatocellular carcinoma, we have carried out investigations at an area where both these diseases develop frequently. 1) We studied the incidence of HBsAg and its antibody in 3521 inhabitants of Tomié-Town in the Goto Islands; 5.5% proved to be positive for HBsAg and 22.0% positive for anti-HBs. 2) In the HBsAg-positive group, the incidence of hepatomegaly and abnormal liver function were significantly higher than those in other groups. 3) In relation to HBsAg, anti-HBs and histological abnormalities of the liver, positive HBsAg was detected in one subject with hepatocellular carcinoma associated with liver cirrhosis, six of nine with liver cirrhosis, and four of 11 with chronic hepatitis. A positive antibody was found in one subject with liver cirrhosis and four with chronic hepatitis. Thus, HB virus infections was closely associated with latent liver disease in Tomié-Town.

Clinical significance of serum alpha-fetoprotein in patients with liver cirrhosis.

The clinical significance of serum alpha-fetoprotein (AFP) was studied in 114 patients with liver cirrhosis, who were followed for more than one year. Out of 114 patients, serum HBs antigen (HBsAg) was positive in 49 (43%). The mean follow-up period was 4.3 years (from 1 to 13 years). Serial determinations of AFP by RIA were performed once a month in all patients. The pattern of serial changes of AFP levels could be divided into three characteristic types: 1) fluctuation of AFP levels to greater than 20ng/ml was observed in 39 cases (Type I); 2) a transient rise of AFP levels to greater than 50ng/ml was observed in 20 cases (Type II); 3) in the remaining 55 cases, AFP levels were always less than 20ng/ml. During the follow-up period, liver cell carcinoma (LCC) developed in 29 cases (25%) and hepatic failure in 11 cases (10%). The incidence of LCC was significantly higher in Type I (18/39, 46%) than in Type II (3/20, 15%) and Type III (8/55, 15%). It is concluded that cirrhotic patients with fluctuation of AFP levels should be included in the high-risk group for LCC.

Tumor/serum ratios of alphafetoprotein and radioimmunolocalization of primary liver cancer.

Twenty six patients with primary liver cancer (PLC) were injected with 131I-labeled antibodies to alphafetoprotein (AFP). Each patient was tested by external photoscanning, 24 and 48 hours after injection. Positive imaging in the tumors was obtained in only 50 percent of patients. Serum AFP levels did not correlate with the radioimmunolocalization results. In 15 patients with PLC, concentrations of AFP in the tumors and sera were determined simultaneously. Although tumor AFP concentrations correlated significantly with serum AFP levels, tumor/serum AFP ratios differed greatly among tumors, indicating that there exists a dissociation between the production and the secretion of AFP. In 8 patients in whom the measurement of serum and tumor AFP concentrations and radioimmunolocalization were performed, tumor/serum AFP ratios correlated significantly with the results of radioimmunolocalization. This suggests that a high tumor/serum AFP ratio is important for the detection of PLC by radioimmunolocalization with antibodies to AFP.

[Evaluation of chemotherapy in hepatocellular carcinoma with liver cirrhosis].

The present paper discusses the efficacy of cancer chemotherapy in 46 patients with advanced hepatocellular carcinoma. In most of patients, Adriamycin (20-40 mg) or Mitomycin C (20 mg) was given by one shot injection via the hepatic artery and followed by serial administration of anticancer agents such as 5-fluorouracil (300-750 mg/day), Adriamycin and Toyomycin. Results were as follows: 1) The complete response was not seen. 2) The partial response (more than 50% decrease of the tumor size) was observed in 6 of 46 patients (13%) for 34 to 457 days. 3) Absence of tumor thrombi in the portal vein which was observed by hepatic angiography, CT and ultrasonography, was closely correlated in the partial response and the prolongation of survival time. 4) There were no significant differences with the therapeutic response or survival time among 3 groups (E2: tumor occupation rate (TOR) 20-40%, E3: TOR 40-60%, E4: TOR above 60%). 5) Obstinate abdominal pain and abnormal liver function remarkably were improved during the chemotherapy in 11 of 18 cases (61%), and 6 of 46 cases (13 %), respectively. 6) Major causes of death were hepatic failure (45.7%), gastrointestinal bleeding (30.4%) and intra-abdominal rupture of the tumor (14.7%). 7) As side effect, some extent of hematopoietic suppression was observed in 25% of the patients treated.

Radioimmunodetection of cancer using antibodies to alpha-fetoprotein and carcinoembryonic antigen.

This study reports the use of radiolabeled antibody to AFP and CEA for the detection and localization of AFP- or CEA-producing tumors. Thirty-one patients received 131I-labeled anti-AFP or anti-CEA antibodies. Photoscans were taken at 24 and 48 hours after injection of radioantibodies. In three of six patients with CEA-producing tumors, radioimmunodetection with anti-CEA antibody showed positive scans. In AFP-producing tumors, 7 of 15 patients had positive findings on immunoscintigraphy using polyclonal anti-AFP antibody, and two of nine patients had positive findings when monoclonal antibodies were used. Analysis of radioantibody in the blood after injection showed both complex and free antibody with immunoreactivity in the circulation, and smaller complexes were seen to form after administration of monoclonal antibodies.

Presence of immunoglobulin G in human sera binding to alphafetoprotein.

Sera from patients with collagen diseases, chronic liver diseases and hepatocellular carcinoma were analysed for alphafetoprotein (AFP)-binding protein. The method used was a 125I-labeled AFP-binding assay with goat anti-human immunoglobulin G (IgG) as the precipitating antibody. AFP-binding IgG was found in the sera from one of 34 patients with systemic lupus erythematosus (SLE) and one of nine patients with hepatocellular carcinoma. This protein may interfere with the conventional radioimmunoassay of human AFP, and may be of interest because of its usefulness as a carrier of radioactivity or therapeutic agents to AFP-producing tumors.

Localizing properties of antibody to alphafetoprotein (AFP) to AFP-producing tumor cells.

Studies have been undertaken to clarify the localizing properties of radioactively labeled antibody to AFP using ascitic hepatoma cells which produce AFP. In the in vitro studies, radioantibodies to AFP were incorporated into the AFP-producing cells, even though excess antigen existed in the medium and the antibodies remained on the cell surface for a long time. In the in vivo studies, the radioantibodies to AFP bound to the AFP-producing cells, especially those producing high levels of AFP (AH66). Tumor localization by scintigraphy was clearer with the AH66 hepatoma than with the AH41C hepatoma, which produces less AFP. The presence of large amounts of AFP in the circulation did not prevent the tumor radioimmunodetection. It is suggested that radioantibodies to AFP bind to the AFP-producing cells despite the presence of large amounts of AFP in the circulation or in ascitic fluid.

Serotherapy of AFP-producing tumors with the purified antibody to AFP.

14 patients with hepatocellular carcinoma and one with a hepatoblastoma were given AFP antibodies and changes in serum AFP levels and clinical courses were observed. Anti-human AFP horse IgG was purified by immunoadsorbent column chromatography. 1 mg of the purified antibody was able to bind 100-150 micrograms of AFP. 200-800 mg were given to each patient, according to their AFP levels. Also, conventional anti-cancer agents were administered consecutively. In all cases, the serum AFP levels measured by RIA decreased rapidly to undetectable levels within 24-48 h after the infusion. Then, in seven of 15 cases, the relatively low level was maintained for periods of 10-96 weeks. In the remaining eight cases, the serum AFP level increased again after 1-2 weeks and rose further with clinical deterioration. Since the antibody administered was undetectable in the serum within 2 weeks after its infusion, it is suggested that the AFP antibody had an inhibitory effect on the production or the release of AFP in some patients.