REPORTS ON SEMINARS

(1) Workshop on “Hepatitis B Virus and Primary Hepatocellular Carcinoma”
Under the auspices of the U.S.-Japan Cooperative Cancer Research Program, a workshop on Hepatitis B Virus and Primary Hepatocellular Carcinoma (PHC) was held in Honolulu, Hawaii, on January 29-30, 1987. The workshop was organized by Dr. Tomoyuki Kitagawa from the Japanese Cancer Institute in Tokyo and Dr. W. Thomas London from the Fox Chase Cancer Center in Philadelphia. The workshop was intended as a forum for a small group of scientists (13 altogether) from the two countries to exchange information and to develop new research projects (see list of participants at end of report).
PHC is primarily a disease of adult men, usually occurring in the fourth through sixth decades of life. Worldwide, about 80% of PHC cases are thought to be etiologically associated with chronic hepatitis B virus (HBV) infection. Most of those HBV infections were probably initiated in the perinatal period or in the first few years of life. Thus, the carcinogenic process must be carried out over several decades. The events which occur over those years that are critical for the development of a tumor are not understood. Much of the workshop was devoted to the discussion of factors and mechanisms which may be necessary for hepatocarcinogenesis.
Dr London reviewed the epidemiology of HBV and PHC, and pointed out that over their lifetimes about 50% of hepatitis B carriers develop liver cancer or cirrhosis; most of the other 50% remain clinically well (although they may have chronic persistent hepatitis or other mild lesions demonstrated histopathologically in their livers). Little is known about the factors that influence the development and rate of progression of serious liver disease. Persistence of viral replication appears to be an important factor. Circumstances which influence the persistence of HBV replication are, therefore, of interest.
Dr. London suggested that iron nutrition and metabolism may affect viral replication and the pathogenesis of PHC. Epidemiologic studies of patients treated with chronic dialysis had generated a model of the role of iron in HBV infection. The model states that iron is required for HBV replication; HBV replication leads to accumulation of iron and increased ferritin in infected hepatocytes and increased ferritin in serum. Hence, cells richer in iron are more likely to replicate virus, and individuals with increased iron stores as denoted by higher ferritin levels are more likely to develop chronic HBV infections, liver damage and PHC. Dr. London cited recent studies showing that, in Taiwan, chronic carriers of HBV with higher ferritin levels were at increased risk of developing PHC, and, in Korea, patients with chronic liver disease and serum ferritin levels persistently greater than 300 ng/ml were at a fourfold greater risk of PHC than similar patients with ferritin levels below 300 ng/ml.
Another factor which may influence viral replication is tuberculosis infection. In three disparate populations (southeast Asian refugees in Philadelphia, native Americans in Alaska and Chinese in Shanghai) positive tuberculin skin tests in HBV carriers are associated with hepatitis B e antigen (HBeAg) negativity. London speculated that M. Tuberculosis infection may stimulate T cells or macrophages to release a factor which inhibits viral replication Bacillus Calmette-Guerin (BCG) could have a similar effect and, hence, might have a therapeutic role in certain carriers.
Dr. Kitagawa introduced the view that initiation-promotion mechanisms derived from experimental hepatocarcinogenesis may be applicable to hepatocarcinogenesis in humans. He reported that initiation-promotion is operative in chemically induced liver tumors both in vivo and in vitro. Initiation requires cell division, and cell division itself has a promoting effect Initiation requires cell division, and cell division itself has a promoting effect. Initiated hepatocytes express altered enzyme patterns (gamma glutanyl transpeptidase, placental type glutathione transferase, etc.). Thus, the number of initiated cells can be enumerated by counting the number of enzyme altered islands resulting form the clonal proliferation of enzyme altered cells. Initiated cells differ in their potential to progress to cancer; only one out of several thousand enzyme altered islands gives rise to a cancer. Chemical carcinogens differ in the spectrum of enzyme altered cells they initiate and, therefore, in the overall risk of inducing a tumor. In chemically induced tumors, carcinogenesis appears to proceed through multiple stages. Initiated (or enzyme altered) cells are resistant to the toxicity of chemicals and the promotion of proliferation of enzyme altered cells by death of unaltered cells may be an important component of carcinogenesis. Dr. Kitagawa viewed the increase in cell replication, which is the reparative part of chronic hepatitis and cirrhosis as both accelerating initiation by carcinogens other than HBV and as exerting a promotional influence on carcinogenesis.
Dr. Harris also emphasized the role chemical carcinogens might play in HBV associated hepatocellular carcinoma. He particularly stressed the possibility that aflatoxin B₁ (AFB₁), either in combination with HBV or independently, might be an important etiologic factor AFB₁ is a procarcinogen which must be activated by mixed function oxidases to an electrophilic metabolite before it can induce carcinogenic effects. Greater than tenfold interindividual differences are observed using in vitro models with cultured human hepatocytes. These differences may relate to oncogenic susceptibility to AFB₁. The mechanisms of action of AFB₁, with or without integrated HBV DNA, may involve activation of cellular oncogenes or inactivation of tumor suppressor genes.
Dr. Harris also discussed recently developed methods for assaying AFB₁ in humans. Although measurement of AFB₁ DNA adducts may be the most relevant to hepatocarcinogenesis, development of assays of sufficient sensitivity has been a problem. AFB₁-albumin or AFB₁-hemoglobin assays may be more practical and more useful for epidemiological studies.
Dr. Kojiro reviewed the relationships between HBV, alcohol, schistosomiasis and thorotrast in the etiology of PHC, using data from 166 consecutive autopsied cases at Kurume University Hospital. Two thirds of these cases (108) occurred in patients with some degree of alcohol intake; 57 of the 108 consumed more than 80 g of ethanol per day for 10 years. A series of 125 surgically resected cases showed a very similar relationship to alcohol intake. Twenty-one to twenty-four percent of cases were hepatitis B surface antigen positive, regardless of alcohol intake. Drinking may accelerate the occurrence of PHC in carriers; the average age of onset was 56 years in heavy drinkers and 62 in non-drinkers. Pure micronodular (alcoholic) cirrhosis was never seen in patients with PHC; there was always a macronodular component, implying that binge drinking or chronic viral hepatitis is usually involved
Schistosoma japonica, which continues to be prevalent in some parts of Japan, may be an etiologic factor in some cases of PHC. Among 4,886 autopsies conducted over the past 20 years, 229 revealed chronic schistosomiasis japonica (CSJ). Of these 229 cases of CSJ, 26% had PHC. Among 399 autopsies of individuals of similar age and sex without CSJ, 8.5% had PHC, a significant difference. The prevalence of HBsAg in CSJ associated cases (26%) was about the same as in other cases of PHC. Morphologically, about two thirds of the CSJ associated PHC cases showed varying degrees of nonschistosomal hepatic changes, including macronodular or mixed micro- and macronodular cirrhosis. Thus, CSJ may contribute to the pathogenesis of PHC but usually is not, by itself, a cause of PHC.
Thorotrast, which can cause atrophy and fibrosis of the liver and may also cause cholangiocarcinomas and angiosarcomas, probably is not a cause of PHC. Dr. Kojiro found only 16 PHCS among 102 thorotrast associated malignancies collected from institutes in various parts of Japan. HBsAg was not present in these 16 cases.
In the review of the entire series of PHC cases, the strongest common denominator was the presence of macronodular or mixed micro- and macronodular cirrhosis. Thus, liver cell regeneration, which is a necessary component of cirrhosis, may be required for the development of PHC.
Dr. Mason discussed early events in the replication of duck hepatitis B virus (DHBV), a virus which appears to cause PHC in ducks in China and Japan but has not been associated with liver tumors in the United States. (This issue was addressed later in the workshop by Dr. Omata). DHBV replicates via a complex process involving reverse transcription of an RNA pregenome. It had been assumed that many of the details of the replication cycle of DHBV and related (hepadna) viruses would have been quickly deduced by analogy to the well-studied retrovirus family. This analogy has not turned out to be the case. Dr. Mason’s laboratory and other workers have found that synthesis of DHBV DNA minus strands is primed by a protein rather than a tRNA, and this result has been further supported by recent observations indicating attachment of the protein to the 5’ terminal nucleotide of the minus strand. Plus strand priming also appears to differ from that of retroviruses. An 18 to 19 base oligoribonucleotide corresponding to the 5’ terminus of the pregenome which is encoded downstream from the plus strand initiation site is probably translocated to prime the plus strand. The primers may also block plus strand completion. Supercoiled viral DNA in infected cells lacks RNA, indicating that the primer must be removed and the plus strand completed during initiation of infection. How the replication cycle occurs may be revealed by exploiting the cultured duck hepatocyte system and specific DNA polymerase inhibitors.
Dr. Matsubara described another system for studying the replication of hepadna viruses. He and his coworkers transfected a human hepatocellular carcinoma cell line, Huh6 (which does not contain integrated HBV DNA) with a recombinant DNA molecule consisting of three tandemly arranged HBV genomes and a neomycin resistant gene. The clone which produced the largest amount of HBsAg was studied in detail. The clone released HBeAg, HBV DNA and Dane particles in the medium, accumulated core particles with single or partially double stranded HBV DNA molecules intracellularly, and produced 3.5 kb, 2.4 kb and 2.1 kb transcripts, the largest transcript corresponding to the pregenome. Although these cells are low producers (5-10 particles per cell), they have produced HBV DNA and Dane-like particles for over two years. They convincingly demonstrate that an integrated HBV genome can act as a template for viral gene expression and replication.
Since 1981, many investigators have shown that integrated HBV DNA sequences are found in almost all cases of PHC with HBsAg in their serum. There has been a controversy about the presence of integrated HBV DNA in HBSAg^- cases. Dr. Mitamura discussed a series of PHC patients from Tokyo in which integrated HBV DNA was detected in tumor tissue in 21 of 23 HBSAg⁺ cases but in none of 30 HBSAg^- cases, 20 of whom were anti-HBs and/or anti-HBc⁺. Clonally integrated HBV DNA was identified in non-neoplastic liver tissue in two cases.
The “hit and run” theory of the role of integration and deletion of HBV DNA in hepatocarcinogenesis was discussed by Dr. Rogler. He suggested that HBV DNA may integrate frequently in cellular DNA and that viral sequences may be excised by the repair enzyme, topoisomerase I. This repair mechanism may result in deletion of chromosomal DNA sequences and translocations. Dr. Rogler’s group identified a cellular DNA deletion associated with an HBV integration localized to chromosome 11p13. Using several gene probes to identify restriction fragment length polymorphisms (RFLPs), they detected the loss of alleles from chromosome 11p in six of 15 PHCs (two at 11p13 and four at 11p15), and from chromosome 13q in six cases. Some tumors involved loss of alleles in both 11p and 13q; others involved one or the other chromosome. Loss of alleles was not detected in several other chromosomes.
Insulin-like growth factor (IGF-II) is located on chromosome 11p. Dr. Rogler proposed that IGF-II may be deregulated by the genetic alterations of chromosome 11p and may promote the growth of preneoplastic nodules into PHCs. In support of this hypothesis, Rogler reported that IGF-II is actively transcribed in many human PHC cell lines, in a majority of woodchuck PHCs and in all preneoplastic nodules from woodchucks chronically infected with woodchuck hepatitis virus (WHV). In the woodchuck tumors with the highest levels of IGF-II transcripts, a new 3.2 kb transcript was seen. This transcript was also seen in all of the preneoplastic nodules but not in normal liver tissues. Furthermore, an inverse relationship of IGF-II transcription and woodchuck hepatitis virus RNA was observed. Tumors with high levels of IGF-II transcripts, and the 3.2 kb transcripts had little or no WHV RNA; tumors with high levels of WHV RNA had no evidence of IGF-II activation. If IGF-II expression is normally increased in immature hepatocytes, then Dr. Rogler’s observations support the hypothesis proposed several years ago by Drs. London and Blumberg that HBV replicates in mature but not less differentiated hepatocytes.
Dr. Hino, who has worked in the laboratories of both Drs. Kitagawa and Rogler, addressed the question of whether integrated HBV DNA, by itself, can be a complete carcinogenic factor. Thus far, HBV has not been demonstrated to contain a viral oncogene. Nor does the promoter insertion mechanism appear operative because a common integration site in host chromosomal DNA has not been identified. As noted above, certain chromosomal rearrangements secondary to HBV DNA integration may be involved. Dr. Hino is testing this hypothesis by creating transgenic mice with different patterns of HBV DNA integration.
Dr. Matsubara presented a potentially very important finding of a new oncogene, named lca, derived from PHC tissues. He and his colleagues transfected NIH 3T3 cells with DNA from 15 PHCs. Very few transformants were observed. Using Alu probes, they found that two foci from two tumors gave similar restriction patterns whereas the transformed foci from three other tumors produced smears. Secondary transfections with the DNA from one of the Alu linked transformants yielded a high frequency of transformation. Further experiments showed that lca is about 10 kb in length, is located on human chromosome 2 and is not linked to HBV DNA. The lca isolated from normal tissue is not transforming but has the same restriction pattern as the transforming lca derived from tumor tissue. Nucleotide sequencing studies are in progress, comparing normal with transforming lca to test the hypothesis that a point mutation in lca is responsible for the transforming property.
Several investigators have suggested that the product of the HBV X gene may be involved in hepatocarcinogenesis. Dr. Moriarity reported on the presence of antibodies in human sera to synthetic peptides corresponding to parts of the X gene. Sera from 86 percent of 21 PHC patients reacted with one or more of four such peptides whereas about 30 percent of sera from cirrhosis or chronic active hepatitis patients reacted with these antibodies. A difficulty in interpreting these observations is that rabbit antibodies to these peptides often cross-react with normal tissue components. Thus, whether the antibodies in the human sera were elicited by viral or host antigens is not certain.
Research on hepatitis B and liver cancer has been slowed by the lack of liver cell lines and the difficulty in culturing primary hepatocytes. In recent years, several investigators have established hepatocellular cancer cell lines with and without integrated HBV DNA sequences. Dr. Murakami and his colleagues have succeeded in deriving five cell lines from PHCs which developed in woodchucks chronically infected with WHV. Four of these five had integrated WHV DNA in the primary tumors plus viral replicative intermediates. Cell lines were established by serially transplanting the tumors in nude mice followed by tissue culture. Only integrated WHV DNA was found in the transplanted tumors and cell lines; none of the integrated sequences included a complete genome.
Recently, Dr. Murakami’s group has focused its attention on a small integrated WHV DNA sequence which lacked any intact viral genes but contained a small (33 bp) region conserved in all mammalian hepadna viruses. Clones of this sequence demonstrated enhancer activity and probably represent the core sequence of the enhancer element. This sequence enhanced transcription of a host mRNA 1.5 kb distant from the WHV DNA sequence. Thus, activation of host genes by integrated hepadna viral enhancing sequences may play a part in the pathogenesis of PHC.
Another approach to the development of a liver cell line was taken by Dr. Kitagawa. He used diethyl nitrosamine (DEN) to produce a non-tumorigenic woodchuck liver cell line. A one-year-old woodchuck without WHV infection was inoculated with 175 mg of DEN over a two-month period. One month after the last injection of DEN, hepatocytes were obtained by perfusing the woodchuck liver with collagenase. Culture of the cells with Phenobarbital gave rise to proliferating foci of epithelial cells from which the cell line, WLC-3, was established.
WLC-3 cells have a cobblestone-like appearance and produce several proteins typical of liver cells. When inoculated subcutaneously into nude mice, they produce tumor-like masses within two weeks but then cease growing and form glandular structures with hepatocellular morphology. The cells do not grow in soft agar. Therefore, WLC-3 cells may be suitable for studies of WHV infection, replication and gene expression and carcinogenesis. For example, DNA sequences from human PHCs or woodchuck PHCs could be used to transfect these cells as an approach to detecting transforming genes.
An important issue with respect to the role of HBV in the etiology and pathogenesis of PHC is whether all hepadna viruses cause liver cancers in their respective species. Of particular concern was the ground squirrel hepatitis virus (GSHV) which did not seem to be associated with PHCs. Dr. Marion and her colleagues have followed colonies of GSHV-infected and uninfected Beechey ground squirrels for a period of six years. Tumors have developed in 21 squirrels; all were over four years of age when the tumors were detected. The predominant type of tumor was PHC, which occurred in 9 of 35 GSHV-bearing animals and in 2 of 32 squirrels with antibody to the virus. Nine of 16 GSHV carriers over four years of age had HCC at necropsy. None of 31 animals free of GSHV markers developed PHC; 12 such animals were over four years of age. Integrated GSHV DNA was found in PHC DNA in three of five carriers of GSHV. Thus, chronic GSHV infection in ground squirrels is similar to persistent infection with the other mammalian hepadna viruses (HBV, WHV) in humans and woodchucks in that it is associated with a very high risk of PHC.
Dr. Omata and his coworkers are concerned with a similar problem; that is, whether chronic infection with DHBV causes liver cancer in ducks. Chinese investigators from mainland China have reported a high incidence of liver tumors in ducks. In recent years it has become apparent that the duck hepatitis B virus (DHBV) is common in Chinese ducks and may have originated in China. Western investigators assumed that the host in China for DHBV was the large, white Peking duck. Dr. Omata, however, found that the more common domestic duck in China was a smaller brown bird. He compared the prevalence of DHBV DNA in the serum of brown and white ducks in Qidong County in China and found that almost 50% of the brown ducks were viremic, compared with only two percent of white ducks. Autopsies showed that the livers of Chinese ducks were commonly affected by chronic hepatitis and PHC, whereas the livers of white ducks were usually normal.
Dr. Omata used virus from Chinese brown ducks to infect one-day old Peking ducklings and established chronic infections in all of the birds. He then divided his flock of chronically infected ducks into four 8roups. One group was treated with aflatoxin B₁, a second group received acetyl-aminofluorine (AAF), a third group was given 3’-methyl-4-dimethyl aminoazobenzene (DAB), and a fourth group was untreated. None of the ducks developed tumors within one year. In the second year, three of three ducks administered aflatoxin developed PHCs. Episomal DHBV DNA, but no integrated DNA, was detected in these tumors. One of the other carcinogen treated birds developed a tumor in the second year, but tumors were not seen in four untreated chronically infected ducks. In the third year, however, three of six untreated birds developed PHCs. Integrated DHBV DNA sequences were found in these ducks.
It is not known, at present, whether DHBV from brown ducks differs from virus isolated from Peking ducks and if it does, if the differences are responsible for the apparent greater pathogenicity observed by Dr. Omata. In any event, these experiments demonstrate that aflatoxin can accelerate the development of PHCs in DHBV infected ducks, but DHBV infection alone is sufficient to produce liver cancer.
The participants in this workshop contributed to an improved characterization of the events which occur between the time of infection with HBV or exposure to a chemical carcinogen and the ultimate development of a liver cancer. Initiation and promotion, terms which have been very helpful in describing experimental carcinogenesis and which have parallels in the pathogenesis of HBV associated PHCs, do not deal adequately with the range of events involved in hepatocarcinogenesis in humans. Further research is needed to delineate the early molecular events caused by aflatoxins and HBV infection, the roles of chemicals, alcohol, chronic HBV infection, the immune system, and nutrients (e.g., iron) in stimulating division of hepatocytes, and the effects on host hepatocytes of integration of HBV DNA sequences. Finally, a genetic change is required to stably transform hepatocytes into neoplastic cells. Several genetic mechanisms have been proposed, including activation of a proto-oncogene, inactivation of tumor suppressor genes (recessive oncogenesis) and chromosome aberrations causing activation of growth promoters, The participants believed that the new systems and techniques described at this workshop could lead to rapid progress in understanding the components of hepatocarcinogenesis in humans.

(2) Symposium on “Genetics of Human Cancer”
Cancer genetics was the subject of a U.S.-Japan Symposium in Los Angeles, California on March 23-25, 1987. The symposium was sponsored by the Interdisciplinary Group of the U.S.-Japan Cooperative Cancer Research Program and was organized by Drs. Hiraku Takebe (Kyoto University) and Frederick P. Li (National Cancer Institute).
Retinoblasoma
Dr. Ei Matsunaga (National Institute of Genetics) began with a presentation of some epidemiologic aspects of retinoblastoma. The incidence of Rb shows some variation worldwide, and appears to be higher in Japan (Kanagawa Prefecture) than in the United States. Retinoblastoma occurs in a hereditary form (up to 40% of cases) and a non-hereditary (sporadic) form. Occurrence of these two forms can be explained by the 2-mutation model of Dr. Knudson. Work on the cytogenetics and molecular biology of retinoblastoma has localized the Rb gene to 13q14. Homozygosity for the mutant allele is a prerequisite for the genesis of this tumor, irrespective of heritability. Environmental factors associated with germinal or somatic mutations leading to Rb were sought in studies of seasonal variation in the months of births and parental ages at birth of Rb patients. No statistically significant deviation was found when compared with controls.
Dr. Masao Sasaki (Kyoto University) continued by discussing the cytogenetics of retinoblastoma. He reported that among 191 retinoblastoma cases, 13 had constitutional chromosome changes, primarily deletions of 13q14, but also translocations involving this region. The skin fibroblasts from familial retinoblastoma patients were susceptible to transformation by murine sarcoma virus. Cells from 13q- patients showed the same susceptibility as normal controls and sporadic unilateral cases. No clear evidence of radiation sensitivity was detected in these fibroblasts. Karyotypes of 37 primary retinoblastoma tumors were studied to investigate cytogenetic changes other than 13q14 alterations. An extra 1q (dup 1q) or iso (6p) were the commonest changes. These observations suggest that the dup (1q) or iso (6p) is an early event of the malignant phenotype.
Dr. William F. Benedict (Children’s Hospital of Los Angeles) traced the efforts to localize, clone and sequence the retinoblastoma gene. In 1980, the gene for esterase D was shown to be in close proximity to the Rb gene on 13q14. Several restriction fragment length polymorphisms (RFLPs) closer to the Rb gene were identified and used to demonstrate loss of heterozygosity in the majority of Rb tumor cells. He described his own work in cloning the Rb gene, including the 5’ portion, and his recent success in sequencing the gene. When Rb tissue was examined with the new probe, internal deletions were demonstrated in the majority of tumors. In all specimens examined, the message was lost, reduced in intensity or abnormal. Likewise, osteosarcoma, fibrosarcoma, and several other individual tumors also showed internal deletions of the retinoblastoma gene. Dr. Benedict’s work provides the first verification at the molecular level of the first and second hits in Dr. Knudson’s 2-mutation model for retinoblastoma.
Renal Tumors
Dr. Yoshiyuki Hanawa (Toho University) discussed the epidemiology of Wilms’ tumor in Japan. He presented data of the Japan Children’s Cancer Registry (JCCR) spanning the period 1969-1984, the Kanagawa Children’s Cancer Registry (KCCR), and the Kanagawa Birth Defects Monitoring Program (KAMP). Wilms’ tumor comprised 4.5% of childhood cancers recorded in the JCCR, a frequency similar to that in most countries. Wilms’ tumor occurred at slightly younger ages in Japan than in the United States or Great Britain. The annual incidence rate of Wilms’ tumor in Japan was 3.8 per million children under 15 years of age, which was low compared to that of other countries. The overall frequency of congenital anomalies associated with Wilms’ tumor was similar in Japan and the United States. However, anomalies of the central nervous system may be more frequent in Japanese children when compared with data from the National Wilms’ Tumor Study Group (NWTS). Further study of racial differences in the incidence of Wilms’ tumor is needed.
Dr. Alfred G. Knudson (Fox Chase Cancer Center) further examined the association of Wilms’ tumor with congenital anomalies. In the aniridia-Wilms’ tumor association, the mechanism appears analogous to that in retinoblastoma, i.e., a chromosome deletion. In Beckwith-Wiedeman syndrome, a consitutional triplication of distal 13p has been found in some cases. However, how a perturbation of distal 13p leads to an increased Wilms’ tumor risk is unclear. He described the genes for retinoblastoma and Wilms’ tumor as examples of anti-oncogenes, loss of which is associated with tumor development. Dr. Knudson also discussed renal cell carcinomas. Known cytogenetic changes in this tumor include a translocation, t(3;8), in a family with multiple affected members, and reports of sporadic renal cancers with chromosome 3 abnormalities. In studies of rats from Norway with dominantly inherited renal carcinoma, increasing numbers of tumors can be induced with an increased radiation therapy dose. In breeding studies, approximately 25% of the litter dies at 9 days of fetal development, but at autopsy no tumors were found. The lethal event may be due to homozygosity of the gene. The rat model may prove important to understanding renal carcinoma in humans.
Colorectal Cancer
Dr. Raju Chaganti (Memorial Sloan-Kettering Cancer Center) presented an epidemiologic analysis undertaken at the Center to evaluate the hereditary component of colorectal cancer not associated with polyposis coli. A family history questionnaire was administered to 458 patients admitted to the Center over a period of nearly three years. The questionnaire was returned by 236 patients. Of the 182 colon cancers reported in first degree relatives records were available for 25. The accuracy rate was 95% among the 25. Comparison of observed and expected cancers in first degree relatives confirmed a three-fold risk of colon cancer in these family members. Data also revealed at least a five-fold colon cancer risk associated with probands under age 40, with right-sided lesions, and with multiple primary cancers.
Dr. Joji Utsunomiya (Hyogo College) presented the heterogeneity of the polyposis syndromes. There are histopathologic variation of the polyps and heterogeneity of extra gastrointestinal lesions. He classified gastrointestinal polyposis into three main histopathologic types: l) adenomatosis coli, 2) Peutz-Jeghers syndrome and 3) juvenile polyposis coli. He further subdivided adenomatous coli into four classes on the basis of extra-gastrointestinal lesions: l(familial polyposis coli (FPC), 2) Gardner syndrome, 3) Turcot syndrome and 4) Zanca syndrome. The numbers of adenomas in the profuse type exceeds 5,000 polyps in total and in sparse type is less than 500 polyps. Gastric cancer has been seen in 3% of polyposis survivors in Japan. Gastric polyps were identified in about 70% of adenomatous coli cases. The evidence suggests some environmental modification in the expression of the adenomatous coli gene.
Dr. Takehiko Sasazuki (Kyushu University) presented molecular and genetic analyses of familial polyposis coli. The incidence of FPC in Japan is estimated to be 1 per 20,000 births, with a penetrance of 93%. Studies at the molecular level revealed: l) no association between FPC and HLA on the basis of linkage analysis; 2) reaction with monoclonal antibodies in some adenomas and adenocarcinomas from the patients with FPC; 3) increased c-myc gene expression in adenocarcinoma; 4) c-myc amplification in adenocarcinoma from 1 of 18 FPC patients; 5) absence of transforming activity with DNA from FPC patients transfected into NIH/3T3 cells; 6) as recipient cells, fibroblasts from FPC patients were not transformed by transfection of activated ras and myc genes.
Mechanisms of Carcinogenesis
Dr. Hiraku Takebe (Kyoto University) opened with a discussion of cancer and mutation in cancer-prone hereditary diseases such as xeroderma pigmentosum (XP), ataxia telangiectasia (AT), Bloom’s syndrome (BS) and Fanconi’s anemia. Quantitative comparison, even in XP fails to confirm similarity in the processes of carcinogenesis and mutagenesis. While XP cells belonging to complementation group A were far more hypermutable by UV than group C, the ages of onset of skin cancer were similar in the two groups. Among other cancer-prone hereditary diseases, hypermutability by mutagenic agents has been shown only for Fanconi’s anemia cells exposed to diepoxybutane, a DNA cross-linking agent. It is still not known whether hereditary predisposition, acting alone or with environmental agents, is responsible for the cancer-proneness in these diseases. Search for the environmental agents with hypermutable action on these cells may lead to identification of the cause of cancer in affected patients. Dr. Kenneth Kraemer (National Cancer Institute) discussed clinical and molecular investigations of xeroderma pigmentosum (XP). He outlined quantitative frequencies of clinical abnormalities seen in XP, as estimated from published descriptions of 830 patients in 297 articles in the medical literature, 1874 to 1982. Clinical features include cutaneous ocular and neurologic abnormalities. The median age of onset of cutaneous findings is between 1 and 2 years of age. Median age of first skin cancer is 8 years, as compared with 60 years in the general population. Basal cell or squamous cell carcinoma of the skin predominate Neurologic abnormalities, including progressive mental deterioration, diminished reflexes and deafness, were reported in 18% overall. A high proportion of Japanese patients was neurologically affected. In laboratory studies, new assays were described which used plasmids as tools to measure DNA repair and mutagenesis in XP cells. Transient expression of a UV-damaged plasmid depends on the competence of cellular repair enzymes. Plasmid survival was reduced in XP cells reflecting repair deficiency, and mutation frequency was increased 100-fold with UV exposure.
Dr. Rufus Day (Cross Cancer Institute) concluded with a discussion of 0-6-methyl quanine repair deficient human tumor cells (called Mer-Minus cells). This system was used to study the relationship between DNA repair defect and susceptiblilty to cancer in syndromes such as XP and Fanconi’s anemia. When a variety of tumor cells were assayed, several were found to be defective in repairing 0-6-methyl guanine. These cells appear to lack a required transferase enzyme. In model cell lines, oncogenic activation appears to be associated with a repair of 0-6-methyl guanine.
Cancer Syndromes
Dr. Shin-ichiro Takai (Osaka Unversity) presented the multiple endocrine neoplasia type 2 syndrome (MEN-2) in Japan. This autosomal dominant disorder is characterized by medullary thyroid carcinoma (MTC) and pheochromocytoma. The MEN-2 is subdivided further into MEN-2A (with parathyroid adenoma) and MEN-2B (mucosal neuromas). The incidence of thyroid carcinoma in Japan is estimated at 1 per 1,000 individuals in the general population. The number of MEN-2A patients in Japan is estimated to be about 600. MEN-2B is rare, with only 15 cases reported in Japan. His cytogenetic studies and linkage analyses fail to place the MEN-2 gene locus at 20p12.2. To date, 12 pairs of tumor and normal DNA samples from MTC patients have been examined for loss of heterozygosity using 15 polymorphic DNA markers. No loss of heterozygosity has been found.
Dr. Stephen Baylin (Johns Hopkins Medical Center) discussed hereditary MTC as a model for the study of human tumor cell heterogeneity. Knowledge is available of the parent cell (thyroid C cells) and its biochemical characteristics in normal differentiation. These properties can be correlated with the progression from pre-neoplastic lesions to macroscopic malignant tumors. Use of calcitonin (CT) as a marker showed that the dopadecarboxylase/CT, ratio and the dopamine oxidase/CT were increased with the development of neoplasia. This was secondary to a decline within cells of calcitonin. He provided a description of the calcitonin gene and its exon 6. The CT gene has been localized to the short arm of chromosome 11.
Dr. Frederick Li (National Cancer Institute) presented studies of families with sarcomas and other neoplasms as a new model for examining inherited cancers. The original observation in 1969 was four families with multiple cases of childhood soft tissue sarcoma in association with brain tumor, leukemia and breast cancer in young relatives. Prospective observation over the next 12 years yielded 16 additional cancers among the four families where less than one case was expected by chance. Additional families have been ascertained with patterns of cancer suggesting autosomal dominant cancer inheritance. Concurrent with epidemiologic studies, cytogenetic analyses of soft tissue sarcomas from sporadic (non-familial) cases have revealed a number of clonal chromosome changes, i.e., synovial sarcomas t(X;18), myxoid liposarcomas t(12;16), and lipomas t(3;12)
Dr. Hiroshi Ogawa (Cancer Center Research Institute) presented data from the Aichi Cancer Registry examining the family history of cancer among cancer patients. The analysis comprised 9,131 cancer patients, 20 years or older at diagnosis, reported to the registry from 1979 to 1981. Approximately 25% of patients had a history of cancer in a parent or sibling. A significant site concordance between study patient and family member was observed for cancers of the breast, colon and rectum, and stomach. These results suggest that cancer of the breast, colon and rectum, and stomach tends to aggregate in families, and that the phenomenon may be explained by genetic backgrounds as well as environmental factors.
Epidemiology of Cancer
Dr. Haruao Sugano (Japanese Foundation for Cancer Research) discussed characteristics of cancer in the Japanese. Cancer trends in Japan are characterized by an increased number of cancer deaths attributable to aging of the population, and a changing cancer pattern associated with a shift in dietary practices. Although stomach cancer remains the most common cancer in Japan (versus lung cancer in western countries), the incidence has decreased. Reported rates of lung, colon, breast and pancreatic cancers are increasing. With increases in fat, meat and calcium consumption cancer patterns are becoming similar to those in the West. In gastric adenocarcimoma, there is a decrease in the ratio of differentiated/undifferentiated lesions in all countries, and a decrease in the ratio of affected males to females to nearly unity in the United States. In the United States, the rate of cancer of the cardia has risen to approximately l/3 of all gastric cancers. The rate of postmenopausal breast carcinoma remains high in women in the United States and low in Japan. However, since 1960, postmenopausal breast cancer has been gradually increasing in Japan, especially in the cities. The rate of prostate cancer has been very low in Japan and among Hawaiian Japanese, but is also gradually increasing. Adult T cell leukemia incidence is high in Kyushu.
Dr. Robert Miller (National Cancer Institute) presented ethnic differences in cancer rates by subsite or subtype. Lymphoma rates vary markedly between U.S. whites and Japanese presumably because of differences in immunologic defenses. There is a reciprocal relationship between the rates of lymphoma (low frequency in Japan) and certain auto-immune diseases (high frequency in Japanese and other Asians). Osteosarcoma has similar rates in various ethnic groups, while Ewing’s sarcoma is rare in non-whites who are apparently resistant to this form of bone cancer. The rate of squamous cell carcinoma of the esophagus is highest in the middle third of the esophagus among black males. Age at diagnosis is younger in blacks than in whites. Adenocarcinoma of the esophagus occurs mostly in white males, primarily in the lower third, presumably due to Barrett’s esophagus which occurs at this site and is rare in non-whites. Adenocarcinoma of the stomach is most common among the Japanese.
Laboratory Methods and Epidemiology
Dr. Chiyoko Sato (Radiation Effects Research Foundation) presented studies of biochemical mutations in the children of atomic bomb survivors. Two groups of children were studied l) children born to parents within 2,000 meters of the hypocenter at the time of bombing (cases), and 2) children born to parents beyond 2,500 meters from the hypocenter or who we not in the city at the time of the bomb (controls). The children were examined with respect to rare variants of 30 blood proteins and 9 erythrocyte enzymes. The results showed similar mutation rates in children of cases and controls
Dr. Robert Sparkes (University of California, Los Angeles) described the use of gene mapping in studies of cancer genetics. This information can have relevance to greater understanding of carcinogenesis and to clinical application. The relationship of specific genes to chromosome breakpoints provides insight into basic mechanisms for the development of specific tumors, such as Burkitt’s lymphoma and chronic myelogenous leukemia. Gene mapping information has made possible the demonstration of somatic cell mechanisms related to tumor development, specifically, formation of the homozygous state in tumors such as retinoblastoma and Wilms’ tumor. Gene mapping information can now be obtained in several ways. The use of somatic cell hybrids made between human and rodent cells continues to provide helpful information; in this situation, phenotypic or DNA probes can be assigned to chromosomes and to parts of chromosomes. An increasingly important technique is in situ hybridization using DNA probes to assign genes to chromosomes and to specific parts of chromosomes. Standard familial genetic linkage analysis continues to be important, especially in genetic counseling. Finally, searching for homozygosity formation for specific gene markers can help identify specific chromosomes and genes on those chromosomes which may be involved in the development of specific cancers.
Dr. Moroi Murata (Chiba Cancer Center) presented data regarding the association between heritable fragile sites and chromosome rearrangements in cancer cells. The population examined included healthy Japanese and patients with leukemia and allied hematologic disorders. As a whole, the incidence of fragile sites among patients was not higher than that found in healthy subjects. However, isolated cases suggest that certain fragile sites may predispose to a few hematologic neoplasms.
Discussion
Dr. Sparkes suggested that the nature of changes in DNA in retinoblastoma in the Caucasian population (e.g., deletion of different parts of the gene) might be compared with that in Japan (or India where the rates are very high). Dr. Knudson thought studies were needed to determine which unilateral retinoblastomas are genetic. Also, he asked whether there is more selection against heritable retinoblastomas than the non-heritable group. The same problems apply to Wilms’ tumor. Are the bilateral tumors genetic? A test is needed and should be at the top of the research agenda after the gene is cloned. Dr. Sparkes said that the FSH gene may be closest to the WT gene, and that one could walk along the chromosome to identify the WT gene starting at the FSH locus.
There is a need to store fibroblasts and Wilms’ tumor (WT) tissue. These should be stored for future DNA studies. Cases with constitutional 11p- are most important. Similar specimens should be sought in Dr. Li’s family under study in which six children had WT, and in Dr. Anna Meadow’s family in which the mother has hemihypertrophy, three children had WT and one of the grandchildren recently born has nephroblastomatosis. Dr. Knudson noted that WT occurred in three genetic settings: a) with aniridia, b) in a hereditary form, and c) in Beckwith-Wiedemann Syndrome (BWS). Tissues from these rare occurrences should also be preserved for future studies.
Dr. Miller asked about the rates of WT, retinoblastoma and neuroblastoma in the Kanagawa Prefecture population-based childhood registry. Dr. Hanawa provided the following data from Dr. K. Nishihara’s registry at the Kanagawa Children’s Medical Center, 1972-1982, under 15 years of age:

Annual Incidence Rate (number affected)
per Million Children, ages 0-14 years

⁺ Hepatoblastoma and hepatocellular carcinoma
^* SEER data added later; population = 12,218,578

Dr. Nishihara issues a report annually on his registry, 1975-1982. Partial reports are in the open literature (in Japanese). Dr. Matsunaga will urge him to publish more of the data in a generally available journal.
Dr. Li said that therapeutic trials for hepatoblastoma in the United States (less than 100 cases/year) will include a brief family history for polyposis, and an eye exam will be made for black spots in the retina indicative of Gardner’s syndrome.
Dr. Miller asked about subacute necrotizing lymphadenitis in children mentioned by Dr. Hanawa. He has observed 10 children at Toho University with the disease, but did not seek etiologic histories. He will ask if the parents were in the medical profession, in view of impression that such people were over-represented in the cluster of cases in Hokkaido, particularly among women 20-34 years of age, Dr. Hanawa’s report of the 10 cases has been submitted for publication. Dr. Miller had mentioned during his presentation that a cluster apparently also existed in Kyung Hee, Korea, according to an abstract on immunologic findings in 15 such cases. Dr. Kokichi Kikuchi in Hokkaido is the physician most interested in this disease. He is seeking an etiologic connection with EBV.
Dr. Miller described the association among four growth abnormalities: cancers (WT hepatoblastoma and adrenocortical tumors singly or as double primaries), congenital hemihypertrophy (too many cells), visceral cytomegaly (Beckwith-Weidemann Syndrome with unusually large cells at birth in the abdominal viscera) and hamartomas (benign developmental tumors). Dr. Sparkes suggested that oncogenes could produce this array of growth disorders Dr. Knudson said it might instead be due to insulin growth factors, which are very active in fetal life.
The screening program in Japan for neonatal neuroblastoma was discussed next. Dr Hanawa said the program covers about 24% of cases, and that 80-85 cases had been found in the last few years. Dr. Knudson asked if n-myc gene amplification in stages III and IV developed later in the natural history of the neoplasm, or if it was present early in the disease. Dr. Hanawa said that even early stage III and IV tumors had n-myc amplifications Dr. Knudson urged that if any stage IV-S tumors were found, to save specimens for future DNA studies to determine, as he proposed some years ago, if it is a hereditary form of the neoplasm.
Ewing’s tumor: The low rates in blacks and Asians, as compared with Caucasians, may be due to certain DNA sequences that are susceptible to 11;22 translocation in Caucasians Perhaps the translocation site can be cloned and examined to see if the gene sequences are dissimilar in Caucasians, blacks, and Asians. Differences might parallel those in African vs American Burkitt’s lymphoma. If growing the tumor cells is difficult, they should be put into somatic cell hybrids.
With regards to breast cancer, Dr. Li asked if neu gene amplification is related to worse survival in the United States than in Japan. The Japanese part of the study could be made at the Cancer Institute in Tokyo.
Polymorphisms in blacks were said to differ from those in whites. Dr. Sato said that RFLP probes provided from the United States, used in studies of whites, differed substantially from polymorphisms in Japanese. Ethnic differences in the frequencies of fragile sites on chromosomes were not thought to be ready for study yet because of uncertainty about their significance and laboratory differences in detecting them.
Dr. Takebe, when asked to comment on the meeting, said “We need genetics on Japanese”. Only 3 of 80 Japanese medical schools have full-time professors of human genetics. There is a need to strengthen genetics in Japan, especially in relation to epidemiology.
With regard to the Li-Fraumeni syndrome, Dr. Miller asked how to seek families with the syndrome in Japan. Dr. Li suggested that young women with breast cancer be asked about soft tissue sarcomas in young family members, or to ask about breast cancer in young female relatives of children with soft tissue sarcomas.
Dr. Knudson stated that the Boston family in which 10 members had renal cell carcinoma should be studied by molecular biologic techniques at the site of the translocation on chromosome 3.
The huge age-peak in testicular cancer among whites, with no peak among blacks or Asians might be studied with respect to isochromosome 12p, which seems to be present in all testicular cancer. These sequences could be cloned and used as a probe in testicular tissue from each of the ethnic groups.



SEMINAR AGENDA AND PARTICIPANTS

(1) JAPAN-UNITED STATES WORKSHOP ON HEPATITIS B VIRUS AND PRIMARY HEPATOCELLULAR CARCINOMA
Honolulu, Hawaii, January 29-30, 1987

AGENDA