REPORTS ON SEMlNARS
(1) Workshop on the Importance of Subtypes of Cancer
A previous workshop of the Interdisciplinary Group on the role of pathologists in cancer epidemiology (San Francisco, March 10-11, 1984) brought out the need to increase attention to subtypes of cancer. As a result, the present workshop was convened October 16-17, 1984, in Honolulu, Hawaii.
Clinically Determined Subtypes
Frederick P. Li (National Cancer Institute, Bethesda, Maryland) described the identification of rare subtypes by clinical means. Such studies are leading to new understanding of cancer biology. He described the 3:8 chromosomal translocation found in a family with 10 cases of renal cell carcinoma and laboratory studies that evaluate the role of the myc oncogene near the breakpoint on chromosome 8. The previously suspected triplication of the myc gene is apparently an artifact, and the breakpoint is now recognized to be more distal to the myc gene locus than previously thought.
Another syndrome involving aplastic anemia and leukemia was mistakenly thought to be Fanconi's anemia. A nurse volunteer who works with Dr. Li noted that the child and other members of the family had cerebellar ataxia (mistaken clinically to be weakness from the anemia). Monosomy 7 in the bone marrow of the patient showed clonal evolution before the occurrence of the leukemia. So, a new syndrome has been defined and a start has been made in understanding the biologic mechanism involved in the associated occurrence of leukemia.
A third syndrome involves soft-tissue sarcomas in more than one child in the family, breast carcinomas among young female relatives, and other cancer in other family members. To date 26 such families have been identified, and in the original four, described in 1969, 10 of 31 surviving family members developed cancer in the next 13 years. Studies of cells exposed in culture to ionizing radiation revealed decreased survival as compared with normal, suggesting a superefficient but error-prone DNA repair mechanism.
Finally, Dr. Li mentioned Wilms' tumor in five cousins whose maternal grandmother had renal cysts and meningioma. In the discussion Dr. Newton called attention to a similar family in Columbus, Ohio, and suggested that future studies be made in collaboration with regard to both families. It was also suggested to him that the Intergroup Rhabdomyosarcoma Study (IRS) consider reporting on the frequency of the soft-tissue sarcoma breast cancer syndrome and on the frequency of breast cancer among relatives of the index cases. Inquiry should also be made to the all Japan Children's Cancer Registry to determine if any instances of this syndrome have been observed. If not, a binational difference in susceptibility may be the explanation. Finally, Dr. Newton agreed to review the histopathology of rhabdomyosarcoma in the syndrome to deterlnine if it fell into a particular subtype. Dr. Li will secure the specimens for this purpose. The same sort of study should be made of rhabdomyosarcoma which occurs excessively in patients with neurofibromatosis.
Subtypes that Have Led to Gene Maps
The next speaker, John J. Mulvihill, M.D. (National Cancer Institute, Bethesda, Maryland) presented the chromosome map of cancer genes that he and Patricia Madigan created in March 1984. This map allows one to relate the cancer gene locus to neighboring noncancer genes, which may be of diagnostic value or nay suggest mechanisms of carcinogenesis, especially with regard to the 18 oncogenes and 21 fragile sites identified to date. The map shows translocations and other chromosomal aberrations associated with lymphoid and/or other forms of cancer. Additions to the map occur constantly, so the present version is already becoming outdated, but is far more informative than its predecessor, constructed by Dr. Mulvihill in 1980. Some chromosomes have multiple loci for cancer genes, and others have none. The "traffic patterns" are well displayed in this formulation (Appendix 1). The discussion brought out the suggestion that the location of genes for growth factors be added.
Subtype Differences in Japanese-Americans as Compared with Caucasians
John W. Berg, M.D. (University of Colorado) had hoped to have detailed information from the SEER Program on cancer in Japanese-Americans for comparison with Caucasians in the U.S., but the data did not arrive in time. He presented the less refined information he had gleaned from preliminary unpublished data specially prepared for him by the SEER Program. Because of the great variety of differences he observed between the two races, it is difficult to summarize his observations except in the summary he provided (Appendix 2). The discussion brought out that what appeared to be an excess of thymomas was probably an excess of germ cell tumors in the thymus (mediastinum), well known to the Japanese pathologists who were present. The data also revealed an excess of pineal tumors among Japanese-Americans as is known to occur in Japan. The possibility of a relationship between the pineal tumor and the germ cell tumors in or near the thymus raised the possibility of a general excess of germ cell tumors along the Japanese as compared with people elsewhere. It is interesting that inquiries in China and Korea have revealed no apparent excess of pineal tumors. Dr. Stemmermann said that virtually all patients in Hawaii with pineal tumors were of Japanese descent. It was suggested that this observation be published, but Dr. Stemmermann said that he did not have access to the cases. Interpretation is that the predisposition is inherent and does not change with migration from Japan to the United States. A question as to which cancer sites show the biggest differences by subtype between the United States and Japan revealed that many did, including lymphomas, brain tumors, bladder tumor, and leukemias.
Grant N. Stemmermann, M.D. (Kuakini Hospital, Honolulu) discussed histologic and subtype differences among common tumors in Hawaiian Japanese on the basis of his experience as a pathologist. A summary of his findings is given in Appendix 3. Among these is the observation that, with regard to stomach cancer, no differences were found in nutrients in cases versus controls or in serum vitamin levels. Better survival is observed for intestinal type tumors, especially among males. One or more polyps in the colon were found in 60% of autopsies of Japanese-Americans. No comparison with whites could be made because the series studied by Dr. Stemmermann came from a hospital that serves only Japanese. Polyps in the descending colon were more frequent among males. The lower the serum cholesterol and the lower the fat intake, the lower was the frequency of colon cancer, especially on the right side. With regard to breast cancer, subtypes among Japanese-Americans were similar to those in Japan and different from those in Caucasian women, who have a higher proportion of in situ cancer. The Japanese have a greater frequency of lymphocytic infiltration. In response to a question, Dr. Stemmemann stated that fibrous tissue is more common in noncancerous breast tissue among Japanese as compared with Caucasians. The question arose because fibrous tissue was noticed to be greater among Japanese in Hiroshima and Nagasaki who had received 50 rad or more from exposure to the atomic bombs than in the U.S. women, and it was not known if the findings in Japanese were naturally occurring or due to the radiation exposure. It is a natural occurrence. Dr. Stemmermann later sent a letter concerning a study that could be made of subtypes of lung cancer in relation to radiation exposure (Appendix 4).
Lymphomas
J. N. P. Davies, M.D. (Albany, New York), at the outset of his presentation, asked in how many ways can cancer be subtyped, when should such refinements be made, and to what purpose? He stated that no contribution to the knowledge of Hodgkin's disease had come from studies of its subtypes and that such study was not likely to help in understanding its etiology. Studies of Hodgkin's disease in Albany, New York, with his coinvestigator, Nicholas J. Vianna, M.D., revealed no insights through the study of subtypes but did show some remarkable instances of unaffected persons closely associated with a series of people who developed Hodgkin's disease. One woman had close associations with at least a half dozen people who later developed Hodgkin's disease (Table 2, Contact C, Vianna NJ et al, Ann Intern Med 77:169-180, 1972). From these observations, the investigators concluded that Hodgkin's disease may be transmitted by an unaffected carrier. Dr. Davies also stressed the potential importance of interacting environmental agents in affecting the subtype, such as malaria and Burkitt's lymphoma or HTLV and AIDS.
Rhabdomyosarcoma
Subtypes of rhabdomyosarcoma have been under intensive study by William A. Newton, Jr., M.D. (Columbus). Rhabdomyosarcoma may be classified clinically as Groups I-IV depending on the extent of the disease and may be separately classified according to tissue patterns as embryonal, alveolar, pleomorphic, or botryoid; or in the current study as classical alveolar, not classical alveolar, or a mixed type. The cytohistology of the 378 patients in IRS I was used to simplify the classification, and the 401 patients in IRS II were evaluated in a prospective study concerning prognosis based on cytohistologic type. It was found that only three of the 43 cytohistologic variables influenced survival. According to the new classification, rhabdomyosarcoma may be classified as monomorphous round cell type, mixed type with anaplasia, or mixed type without anaplasia. The first two more often have unfavorable outcomes than the mixed type with anaplasia:
| Old histologic classification |
Embryonal, alveolar, botryoid, pleomorphic |
|
|
| New histologic classification |
Classical alveolar*, not classical alveolar, mixed pattern |
|
|
| Cytohistologic classification |
Monomorphous round cell type, mixed anaplastic, not anaplastic* |
*Favorable prognosis
Intermediate prognosis
General Discussion
Dr. Miller led the discussion first to a consideration of the various ways in which subtypes of cancers could be described:
Morphologic
- Gross anatomy, with respect to location in the bowel or stomach, the natural history, and metastatic patterns
- Conventional tissue patterns (for example, alveolar in rhabdomyosarcoma)
- Cytohistologic, which concerns fine details within the cells
- Histochemical (e.g., glycogen present or absent)
- Antigen identification, including cell surface markers
- Ultrastructural, used mainly for confirmation or to support a diagnosis
- Cell products and chemical receptors, such as mucous, pepsinogen, or hormonal
- Biochemical
- Chromosomal studies and DNA
Clinical
- Familial versus sporadic
- Response to therapy
- Host susceptibility
Epidemiology
- Demography
- Ethnic differences
- Chemoprevention
In brief, subclassification extends from studies of populations to studies of DNA, a surprising realization.
Dr. Miller then summarized ideas for new research which are embodied in the text above and in the summary for the Japanese papers being prepared by Dr. Sugano.
It is well known that cancer of a certain organ can be divided into two or three (or more) subtypes by histological features. For example, gastric cancer is classified into two types, intestinal and diffuse. These two carcinomas are different not only in histological features, but also in their morphogenesis, clinical behaviors, and survival rates. Further, these two types should be different in their etiologies. It would be worthwhile to approach cancer etiology and biology in a certain cancer through analysis of subtypes. According to recent biochemical and immunological advances, tumors are also subclassified into various cancer phenotypes. This workshop had two aspects; one was the organ site, in which subtypes of tumors of the stomach, lung, and breast were discussed. Another was tumor markers, in which subtypes by means of cell atypism, monoclonal antibody, hormone, and hormone receptors were introduced.
Tumor Cell Atypism and Automated Cytology
Noboru Tanaka (Chiba Cancer Research Institute, Chiba) explained recent developments in automated cytology. There are two systems in automated cytology-the flow system and the image analysis system. The former has been extensively studied in the U.S. and the latter in Japan. Dr. Tanaka has developed a system, called "CYBEST model 4," which as various optical systems and requires 3 minutes per specimen for the final assessment.
Squamous cell carcinoma of the cervix is classified into three subtypes--keratinizing, nonkeratinizing large cell, and nonkeratinizing small cell--and prognosis of patients with cervical carcinoma corresponds to the subtypes. The prognosis for keratinizing type is favorable; for nonkeratinizing small cell type it is poor; and for nonkeratinizing large cell type the prognosis falls between these two. Accuracy of cytologic evaluation of subtypes appeared to be specific, 97.94% for each subtype. According to his machine, morphometric differences among these subtypes were fairly good. Finally, Dr. Tanaka pointed out the involvement of human papilloma virus and herpes simplex type 2 in the lesions and neoplasia of the cervix.
The question arose whether CYBEST was practically applicable for diagnosis. Dr. Tanaka answered that it was still at the test stage, but cytodiagnostic work carried out at present would be automated in the near future.
Basic Versus Variable Subtypes
Haruo Sugano (Cancer Institute, Tokyo) introduced his idea of the basic and variable subtypes of cancer. The basic cancer is a proper cancer of a certain organ, and it is less changeable in its incidence by environmental conditions; on the contrary, the variable one is a cancer whose incidence is more easily modified by environmental changes. From an etiologic point of view, the variable cancer may be strongly related to environmental carcinogens; the basic cancer, however, may be closely related to genetics and/or susceptibility of a host. This idea came from the works on natural history of cancer. Dr. Sugano demonstrated that this concept was applicable to cancers of the stomach, lung, and other organs.
Stomach: Gastric carcinoma is histologically divided into two sub-types--poorly differentiated (diffuse) (PCA) and differentiated (intestinal) (DCA). The former arises from the nonintestinalized gastric mucosa and the latter from the intestinalized mucosa. The time trend data indicate that in both sexes the incidence of DCA decreased but that of PCA is steady. Therefore, in the stomach PCA is classified as a basic cancer while DCA belongs to the variable cancer subtype.
Lung: Fundamentally, squamous cell carcinoma and small cell carcinoma belong to the variable cancer, while adenocarcinoma tends to be the basic cancer; but in many countries adenocarcinoma consists of both basic and variable types.
Breast: Premenopausal cancer is the basic cancer and postmenopausal the variable one.
Colon: Carcinoma in young people tends to be of the basic cancer subtype and cancer in older people to be the variable subtype. Most cancers of the colon in Western countries are considered to be variable cancers.
Lymphoma: The incidence and histology of nodal lymphoma varies by country, whereas lymphoma of the gastrointestinal tract does not vary in this same manner.
Melanoma: There are two kinds of melanoma--sunlight related and sunlight unrelated.
Breast Cancer and Subtypes
Goi Sakamoto (Cancer Institute, Tokyo) discussed the present status of breast cancer in the Japanese in comparison with the Americans. Breast cancer can be divided into two subtypes--premenopausal and postmenopausal. The incidence of premenopausal cancer is not as different by country as is that of postmenopausal cancer; for instance, postmenopausal breast cancer incidence is very high in the United States and very low in Japan. Breast cancer in Japan recently increased rather steeply, especially in postmenopausal women. Dr. Sakamoto stressed the important role of obesity in development of breast cancer, especially in postmenopausal women. In women over 60 years of age, the relative risk for breast cancer is 2.5 (p < 0.01) in obese versus nonobese women. Nutritional conditions in Japan improved rapidly after World War II and body height and weight increased. The obese index also increased, especially in postmenopausal women. Therefore, it was concluded that the increase of postmenopausal breast cancer in Japan was correlated with the fact that post-menopausal women tend to be obese.
Dr. Stemmermann discussed that both pre- and postmenopausal Hawaiian Japanese women were frequently found to have in situ breast carcinoma--1 to 3% in hospital cases but 15% in public survey cases. These figures were over two-fold higher than that of caucasians. Further, Japanese breast cancer was often accompanied by lymphoid infiltration around the primary tumor. In relation to obesity and breast cancer, Dr. Stemmermann pointed out that fat intake was also related to body fat and breast cancer development. In addition, patients with hypertension and diabetes mellitus tend to develop breast cancer. Dr. Davis asked what the effect of the Japanese belt (obi) was in the development of breast cancer. No one could answer. To a question regarding what tumors were westernizing like breast cancer in Japan, Dr. Sakamoto replied that most cancers such as lung, colon, pancreatic, and endometrial cancer have increased recently.
Lung Cancer and Subtypes
Yukio Shimosato (National Cancer Center, Tokyo) discussed subtypes of lung cancer, mainly squamous cell carcinoma and adenocarcinoma, and their morphogenesis in relation to cigarette smoking. Lung cancer is histologically classified into four major types--squamous cell carcinoma, small cell carcinoma, adenocarcinoma, and large cell carcinoma. Among surgically resected lung cancers, adenocarcinoma is the highest in frequency in Japan (46.5%) followed by squamous cell carcinoma (34.9%). The frequency of squamous cell carcinoma and adenocarcinoma varies very much by sex. The male to female ratio in cases with squamous cell carcinoma was 15:1; on the contrary, that in cases with adenocarcinoma was 2:1. Almost all male patients with squamous cell carcinoma were heavy smokers and one-half of females were nonsmokers.
It is well known that most squamous cell carcinoma is of central origin but, according to Dr. Shimosato, squamous cell carcinomas of nonsmokers were mostly peripheral in origin, and in many instances they showed histologically some adenocarcinomatous features as to location and cellular characteristics, such as secretory component. The majority of adenocarcinoma arose in the periphery of the lung. Histologically, it tends to be more highly differentiated in nonsmokers than in smokers. Highly differentiated adenocarcinoma is not associated with smoking, and development of poorly differentiated adenocarcinoma may be related to cigarette smoking. These results may indicate that histology of both squamous cell carcinoma and adenocarcinoma is somewhat different between smokers and nonsmokers.
The origin of small cell carcinoma was discussed and it was asked whether macrophages changed into small cell carcinoma of the lung. Dr. Shimosato answered that he did a comparative study on macrophages and small cell carcinomas using various histo- and cytochemical markers and found no similarities between these two; therefore, he believed that the macrophage origin hypothesis should be discarded. Partial deletion of chromosome 2 in small cell carcinoma was also a subject of discussion. Dr. Shimosato stated that he and his colleagues examined the karyotype of several cases of small cell carcinoma and found various chromosomal anomalies, but partial deletion of chromosome 2 was not specific.
Cancer of the Lung and Stomach: Incidence and Subtypes
Aya Hanai (Center for Adult Disease, Osaka) discussed trends of cancer incidence of the lung and stomach according to histological subtypes in Osaka. Lung cancer incidence increased steeply, and its age-adjusted incidence rates were 35 and 10 for male and female, respectively, in 1981. During the past 9 years, cumulative incidence rates of adenocarcinoma and undifferentiated carcinoma, consisting mostly of small cell carcinoma, in males have increased by 133% and 165%, respectively, whereas no change has been found in squamous cell carcinoma. In females, adenocarcinoma has also increased by 126%, but others showed no change. A case-control study was conducted to investigate the risk of smoking to each subtype of lung cancer. Relative risk of smoking to lung cancer was highest for small cell carcinoma in males followed by squamous cell, large cell, and adenocarcinoma. These all were statistically significant and dose response relation was observed in all subtypes. In females the risk f or adenocarcinoma was calculated as 1.7 and, therefore, statistically insignificant.
During the past 15 years, stomach cancer incidence and mortality decreased 25% and 32%, respectively, in Japan. A more rapid decrease was observed in incidence of the intestinal type carcinoma than that of the diffuse type during 1968-72 and 1973-77. However, the declining rate in the incidence of the intestinal type has become the same as the diffuse type rate during the 1978-81 period.
Dr. Stemmermann discussed that both smoking and alcohol are risk factors for all subtypes of lung cancer in Hawaii. Smoking is a risk factor for stomach cancer of both the intestinal and diffuse types; however, alcohol is not a risk factor for stomach cancer. Dr. Miller asked why stomach cancer of young people progressed so rapidly and also what the role of pregnancy was in the progression of stomach cancer. The effects of hormones were discussed. Dr. Sugano stated that stomach cancer was often positive for an estrogen receptor, but the effects of endocrine therapy (tamoxifen) were uncertain. Further, estrogen inhibits the development of stomach cancer in animal experiments.
Monoclonal Antibodies and Subtypes of B Lymphoma
Kokicki Kikuchi (Sapporo Medical School, Sapporo) discussed monoclonal antibodies as immunological markers of lymphoid cells and subtypes of B-cell lymphoma. Dr. Kikuchi developed 12 different monoclonal antibodies for B cells; some show strong reactions to a B-cell lineage of differentiation and some react only slightly. Therefore, using these monoclonal antibodies, the stage of maturation of B-cell lineage and characteristics of lymphoma can be defined. These antibodies clearly and objectively demonstrated immunohisto- and cytochemical characteristics of infiltrating cells and their topographic distributions in the lymph nodes and lesions of lymphomas as well as lymphoid diseases. Further quantitative analysis of these infiltrating lymphoid cells was available by flow cytometry. These monoclonal antibodies were also helpful in the differential diagnosis of subtypes of B lymphoma, and 45 lymphoma cases were precisely classified into 11 subtypes. In lymphoma, cell surface phenotypes often relate to the nature of lymphoma cells; the identification of sub-types of lymphoid malignancies according to antigen profiles defined by these monoclonal antibodies should be a reflection of the pathological features and clinical courses of patients.
Dr. Li stated that chronic lymphocytic leukemia (CLL) was very infrequent in Japan and that Japanese lymphomas had special features in general. These features may suggest the existence of certain defects of normal lymphoid cells of the Japanese. Is it possible to detect those differences with monoclonal antibodies? Dr. Kikuchi replied that no exact comparative studies had been carried out, but people had thought that there was no difference between Japanese and caucasians. The importance of such a comparative study was discussed.
Tumor Subtype and Hormone Production
Hiroo Imura (Kyoto University, Kyoto) discussed hormone-producing tumors, especially ectopic ones. If one divides tumors into apudoma (tumor derived from cell APUD series) and nonapudoma, 80% of ectopic ACTH-producing tumor accompanying hyperadrenocortism belong to apudoma. Asymptomatic ACTH production is common and, if measured in randomly selected tumors, ACTH is detected in 70% of apudoma and 30% of nonapudoma. Another hormone, calcitonin, also was highly detected in apudoma (apudoma 79% versus nonapudoma 32%). Other various hormones, such as vasopressin, oxytocin, somatostatin, corticotropin-releasing factor (CRF), and others, are produced by apudoma. On the other hand, hCG and hPL are produced mostly by nonapudoma.
To clarify the mechanism of ectopic hormone production of tumor cells and whether the genes coding for such hormones in tumor tissues are different from that in normal tissues, Dr. Imura obtained cDNA clones from mRNA for ACTH and CRF precursors and demonstrated the mRNA coding for these hormones in ectopic hormone-producing tumor by dot hybridization. However, both hormones and mRNAs can be detected in normal tissues other than the tumor tissues. Further, the ACTH gene and CRF gene obtained from tumor tissues are structurally the same as those from the normal tissue. It seems to be important to examine the relationship between hormone-related genes and oncogenes in tumor cells, but there is no direct evidence for a relationship between them. It is known that enhancer is responsible for determining the tissue specificity, so that a set of enhancer-promoter-ACTH genes should be considered when particular tumnors or tissues produce particular hormones. Finally, Dr. Imura mentioned that detection of hormones was useful for the diagnosis of tumors, but it would be limited to ectopic and eutopic hormone-producing tumors.
Hormone Receptors and Subtypes
Junzo Kato (Yamanashi Medical University, Yamanashi) discussed the role of the sex steroid hormone receptor as a marker of tumor responsiveness in hormone-dependent cancer. Receptor analysis is useful for prediction of hormonal responsiveness of the hormone-dependent cancers. In breast cancer, clinical evaluation of hormone receptors has been established; about 60% of breast cancers were positive for estrogen receptor (ER), and about 50% of these ER-positive breast cancers were responsive to endocrine therapy. Cancers with high ER concentrations and/or both ER- and progesterone receptor (PR)-positive cases were much more responsive to endocrine therapy; however, no correlation exists in breast cancer between histopathological subtype and ER or PR concentrations.
Endometrial cancer is also hormone dependent. Its remission rate with antiprogesterone has been reported to be 30-57%. According to Dr. Kato, there was no correlation between subtype of endometrial cancer and ER or PR concentrations. Namely, the concentrations in the poorly, moderately, and well-differentiated cancers were as follows: for ER, 11, 21, and 71, and for PR, 20, 51, and 289, respectively. They were also very high in endometrial hyperplasia. A wide range of receptor values in each group of endometrial cancers may indicate variable hormonal responsiveness to the tumors. Distribution of the receptors in endometrial cancer by grade was as follows: ER+ PR+ in 80% of well-differentiated tumors but ER- PR- in 50% of poorly differentiated tumor.
The relationship between androgen receptor (AR) concentrations and histological subtypes in prostatic cancers was the same as that in endometrial cancers--highly differentiated tumors indicated high AR concentrations and poorly differentiated ones low concentrations. The responsibility for hormone therapy of prostatic tumors is the same as in endometrial cancer; high responders were found among cancers with high AR concentrations and low responders among cancer with low AR concentrations.
Finally, Dr. Kato stated that although the number of cases examined were small, survival rate might not relate to ER, but to PR. As to the evaluation of hormonal sensitization by tamoxifen, induction of PR by tamoxifen was found in not only well-differentiated tumors, but also in low PR tumors. Therefore, a tamoxifen challenge test may be useful as an in vivo tumor sensitivity test.
During the discussion, questions were asked about ER-positive tumors other than breast and endometrial cancers. About 15-20% of ovarian tumors were positive for ER and some cervical tumors were also positive. Twenty percent of gastric cancers in Japan were positive for ER.
This was the first meeting to discuss the relationship between cancer subtypes and etiology and biology of cancer. We wished to open a new approach to etiology--biology as well as diagnosis in human cancers through the study of cancer subtypes.
(2) Workshop on Adult-Type Cancer Under Age 30
New understanding of the etiology and pathogenesis of cancer may come from studying its occurrence at unlikely ages or anatomic sites. With this in mind, a workshop on adult-type cancer in persons under 30 years of age was held in Tokyo on March 11-13, 1985, under the sponsorship of the US-Japan Cooperative Cancer Research Program.
For the United States good estimates of the incidence of cancer by site and histologic type are available through the Surveillance, Epidemiology and End Results (SEER) program of the National Cancer Institute. About 10% of the U.S. population is covered by the registry, and more than 95% of the diagnoses have been histologically studied. Young (NCI) had prepared detailed tables to show these data, by age, sex, and certain subsites for whites, blacks, and Japanese-Americans, 1973-1982. Table 1 shows findings of interest in whites under 30 years of age as compared with the corresponding mortality data for Japan, 1968-1978, for persons under 30 years of age, as presented by Aoki (Nagoya).
Death certificate diagnoses are of variable accuracy and completeness. They are population based, however, and are representative of the cancer experience according to current practice as it relates to making diagnoses and coding them on death certificates. Comparisons between mortality and morbidity will, of course, be better for cancers that are quickly fatal than for those with long survival. The data are given as numbers of cases rather than rates to bring to mind the opportunities for studies of case series. Data from the Nagasaki Tumor Registry, presented by Hoel, Mabuchi, and Davis (Hiroshima) and Ikeda (Nagasaki) will be used in this report to supplement observations concerning individual cancer sites. These data are not yet population based and have other problems in ascertainment. Some data were also presented from the Hiroshima Tissue Registry. (The workshop may have helped focus attention on the potential value of these registries and the need to develop them further.) Data were also available from a variety of other registries, e.g., the All Japan Children's Cancer Registry, the Autopsy Annuals in Japan, and the Bone Tumor Registry in Japan.
A variety of findings deserve additional study, especially the marked differences in the occurrence of specific cancers in the two countries and unexpected clusters within countries.
Oral Cancers
Lip: From the SEER data one can see that there were 29 new cases of cancer of the lip in U.S. white males and four in females. Subsequent investigation showed that among the 10 SEER registries, the cases in males clustered in New Mexico and Utah. The excess was not in the South as might be expected if smokeless tobacco, a carcinogen in vogue among young males, were responsible. Further studies should be made.
Gums and mouth: The SEER data revealed a great female predominance for cancer of the gums and mouth: 32 females versus 13 males. Further study of this series may be etiologically revealing.
Tongue: Cancer of the tongue caused 102 deaths (nine per year) in patients under 30 in Japan and 46 new cases annually in the United States. Because xeroderma pigmentosum (XP), a genetic disorder, predisposes young people to cancer of the tip of the tongue, the location of the neoplasm is of interest and was readily ascertained in the SEER Program. The tip was the most frequently affected location. Of the 31 cases in which the location was specified, 16 were at the tip and another five were in the anterior two-thirds. Examinations of other clinical data should show if any of these patients had XP.
Gastrointestinal Cancers
Esophagus: It is surprising that any cancers of the esophagus were reported in the two series, but five deaths occurred in Japan in 11 years, and an estimated five new cases are reported annually in the United States. Again, a genetic disorder may be involved, for tylosis predisposes to esophageal cancer.
Stomach cancer: The well-known high rate of stomach cancer in Japan was also found among young adults. The SEER Program showed about 50 new cases per year occurred among U.S. whites as compared with about 550 deaths per year under age 30 in Japan. Correa (New Orleans) noted that in Japan diffuse carcinoma predominates in the young, as contrasted with the intestinal type later in life. The diffuse type was thought to vary little in frequency with age, sex, race, migration, or geography, but the frequency is clearly higher in young Japanese than in young people in western countries. This finding indicates a need to reevaluate the conventional wisdom concerning the stability of rates for the diffuse form of gastric cancer and to study its origins in young Japanese.
The female predominance is of interest: 3472 females versus 2483 males in the Japanese death certificate series. By contrast, the SEER data showed a male predominance: 33 to 17.
Correa pointed out that the diffuse type of gastric cancer is more often found in persons with blood type A (35%) versus the intestinal type (22%) and the general population (19%). He suggested a multicenter study to accumulate a substantial number of cases, with pedigrees noted and a search made for genetic markers. The role of genetics seems to be stronger than that for the environment. The high mortality rates indicate that specimens and information can be accumlated rapidly in Japan.
Colon cancer: Stemmermann (Honolulu) reported that among records for Japanese-Americans seen at Kuakini Hospital, he found seven with colon cancer under 30, one with multiple polyposis, two in single polyps, and the other four had rapidly fatal disease. The Hiroshima Tissue Tumor Registry, 1973-82, accessioned 33 cases of colon cancer under age 30. One patient had polyposis coli and three had cancers that arose from isolated polyps. In the summary of data from the Autopsy Annuals in Japan, 1974-82, 64 males and 47 females under 30 were diagnosed as having cancer of the colon or rectum. The distribution of these cases by histologic subtypes and anatomic sites could be compared with those of the previously mentioned small case series.
Utsunomiya (Tokyo) presented data from the Japanese Polyposis Registry System. The frequency of multiple polyposis is about the same in Japan as in England, one in 22,000 births. Of 185 Japanese patients with polyposis coli diagnosed under 30 years, 57 had cancer of the colon. [If there were no bias in ascertainment of cancer cases over those without cancer, it can be calculated from the prevalence rates given by Utsunomiya that in the United States one in eight cases of colon cancer under 30 should have polyposis coli.] Colon cancer is usually mucinous when the onset is early in life. This was in accord with observations at Kuakini Hospital, but in Hiroshima only one of 33 cases under age 30 was specified as mucinous.
Stemmermann and Utsunomiya concluded that there was no evidence of a strong genetic influence on colon cancer under age 30. The SEER cases, however, night be studied for a family history of colon cancer; i.e., familial colon cancer without polyposis. There are about 277 new cases per year under 30 in U.S. whites. In both countries the appendix was the site of colon cancer more often than one would expect from experience in pathology. There were an average of 21 deaths a year in Japan and 97 new cases annually among U.S. whites. Subsequent examination of the SEER data showed that all but 7 cases were carcinoids, and 43 of the 90 were reported in Seattle. An evaluation should now be made to determine if the cluster is real or due to the particular criteria used for diagnosing carcinoids in Seattle.
Stemmermann stressed (1) the importance of studying the frequency with which subsites of the colon are affected by various subtypes with respect to differences in diet and (2) the importance of pathologic confirmation.
Rectal cancer: Cancer of the rectum in young adults was 1.4-fold more common in Japan than in U.S. whites, the annual numbers of cases being 86 and 62, respectively. In contrast, a 2.8-fold excess of colon cancer occurred in U.S. whites as compared with Japanese under 30 years of age.
Pancreatic cancer: Cancer of the pancreas occurred about twice as frequently as a cause of death in Japan as in the SEER series. Study of the 33 cases annually in each country might reveal clues to etiology or mechanism that have eluded detection in the many large studies of older persons with this cancer.
Liver cancer: Hepatoblastoma is a neoplasm of children with a peak soon after birth. Hepatocellular carcinoma, the adult form of liver cancer, has been the subject of several U.S.-Japan workshops and was not dealt with in this one. The 48 new cases in the SEER Program among U.S. whites at 15-29 years contrasted with about 26 cases expected at the corresponding Japanese rate.
Cancer of the gallbladder late in life is substantially more common in Japan than in the United States. The rates under 30 years were low in both countries but were about twice as high in Japan where there were an average of 11 deaths per year.
Respiratory System
Maxillary sinuses: The high rate of maxillary sinus cancer in Japan is well known there. An excess of cases in patients under 30 years of age is suggested by the mortality data, 17 deaths per year, which is about double the number of cases expected on the basis of the SEER data.
Lungs and bronchi: Although lung cancer is far more frequent in the United States than in Japan, no such excess occurs under 30 years of age. Japan averages 66 deaths a year, which is about the number expected according to SEER data. However, in the SEER series carcinoid was the diagnosis in 39 females and 21 males. Study of the series of 60 carcinoids would also be of interest, perhaps in conjunction with the carcinoids of the appendix.
Other cancers of the lungs and bronchi affected 35 females and 40 males, and virtually all are known to be primary in bronchopulmonary tissue. The lack of a sex difference in the United States is of interest. The histories of cases in this series may be informative. For example, information might be obtained to determine if the affected people were smokers who had been exposed to the fallout in schoolrooms from asbestos-sprayed ceilings. In this connection it is interesting to note that among the 75 bronchopulmonary cancer cases, seven males and no females had oat cell tumors, 12 (both sexes) had squamous cell cancer, 18 had adenocarcinoma, and 38 had other types. The family histories might reveal a propensity toward certain of the cell types.
Bone Cancer
The SEER data showed osteosarcoma and Ewing's sarcoma to be equally frequent in males (153 versus 156 cases), but there was a relative deficiency of Ewing's tumor in females (138 versus 81). Miller (Bethesda) had examined these data by skeletal site affected and found that Ewing's sarcoma occurred especially in the axial bones and legs, whereas osteosarcoma occurred in the long bones, particularly of the legs. Namba (Hiroshima), a hematopathologist, believed the anatomic distribution reflected the marrow origins of Ewing's sarcoma as contrasted with the origins of osteosarcoma at centers of bone growth.
Data from the Bone Tumor Registry in Japan, 1972-1982, presented by Furuya (Tokyo), showed a marked deficiency in Ewing's sarcoma relative to osteosarcoma as compared with SEER data:
|
Bone Cancer
|
US Whites
|
Japan
|
|
M
|
F
|
M
|
F
|
| Osteosarcoma |
153
|
138
|
629
|
413
|
| Ewing’s sarcoma |
156
|
81
|
84
|
68
|
| OS /ET |
1.0
|
1.7
|
7.5
|
6.1
|
It will be noted that Japanese males had 7.5 times more osteosarcoma than Ewing's sarcoma, and females had 6.1-fold excess of osteosarcoma versus 1.0 and 1.7 for U.S. white males and females, respectively. A marked deficiency of Ewing's sarcoma in U.S. blacks has previously been recognized. This persists despite their long presence in the United States. Presumably, U.S. whites and blacks have similar childhood exposures, but blacks are rarely affected by Ewing's sarcoma, as if they are genetically resistant.
The Bone Tumor Registry shows that chondrosarcoma is less frequent in Japan than in the United States, as measured against osteosarcoma:
|
Bone Cancer
|
US Whites
|
Japan
|
|
No.
|
|
No.
|
|
| Osteosarcoma |
291
|
77.8
|
1042
|
90.5
|
| Chondrosarcoma |
83
|
22.2
|
111
|
9.6
|
| Total |
374
|
100.0
|
1153
|
100.1
|
Soft tissue sarcomas: Differences in criteria for classifying subtypes of soft tissue cancers make it difficult to compare the two sets of data. One finding of possible interest concerns 63 persons with nerve sheath tumors under 30 years of age registered in the SEER Program. If not already done elsewhere, it would be interesting to determine how many were isolated findings or part of multiple neurofibromatosis, whether recognized at diagnosis as such or not.
Skin: Skin cancer is rare under 30 years of age, except in the genetic disorder xeroderma pigmentosum (XP). Takebe (Kyoto) reported that in a series of 184 Japanese patients with XP, 98 have developed skin cancer. A graph by Kraemer et al. shows how the age of the susceptibility to skin cancer is dramatically reduced by XP. Susceptibility is due to a defect in repair of DNA after exposure to ultraviolet light. The greater the reduction in unscheduled DNA synthesis (UDS), the more severe are the symptoms in XP. In Japan the distribution of cases according to the seven genetic complementation groups and the variant group is different from those in the United States. Europe, and Egypt, where XP is especially prevalent. Of 56 XP families in Japan, 48 are in complementation A or the variant group. Elsewhere there is heavier representation from groups C and D and much less from the variant group. The average age for the onset of skin cancer in group A and those with less than 5% UDS was 9.4 years. The closer the UDS level was to normal, the older was the age at onset of skin cancer.
To date, complementation group F (four families) has been found only in Japan. In Korea the distribution by complementation groups appears to be different from that in Japan, but to date only four cases have been studied. The incidence of XP in Japan is estimated to be one in 100,000, which is higher than in other countries. Takebe said that Bloom's syndrome, a genetic disorder primarily of the Ashkenzic Jews, has been found in nine Japanese. Two have developed cancer, including one with lung carcinoma at 39 years. A special Bloom's syndrome registry developed by German consists of 100 patients, 25 of whom have developed cancer (leukemia, lymphoma, and carcinomas), usually under 30 years of age. These syndromes illustrate how genetic factors can play a role in the development of adult-type cancers at much earlier ages than usual. One might look for formes frustes of genetic disorders in persons with early onset of adult-type cancer.
Melanoma: Melanoma is apparently much less frequent in Japanese than in U.S. whites. The influence of ultraviolet light in its induction may be blocked by skin pigmentation. The SEER data show that in the decade studied, 681 males and 1047 females had melanomas. The well-characterized sex difference in anatomic distribution was found: less involvement of the face and scalp of females, but much more on the extremities than in males.
Kasuga (Tokyo) reported that the Annual of Pathological Autopsy Cases in Japan (APAC), 1958-1982, described 69 melanomas under 30 years of age and 29 from surgical cases seen at the Cancer Institute in Tokyo, 1947-1981. Of particular interest was the finding that 16 of the 69 cases from APAC were of the leptomeninges, a seemingly high frequency. The peak was at about 20 years of age. Under 30 years of age, 23.2% of melanomas were of the meninges as compared with 2.0% of melanomas in persons 30 years of age or older. It would be important to confirm this observation from other data, if possible, and to determine if involvement of the meninges was familiar or associated with other disorders in the patient or family. Melanomas of the mucosa were said to be more frequent in Japan than in western countries.
Genital Cancers
Prostate: Cancer early in life coded to the prostate is likely to be rhabdomyosarcoma. The SEER data show eight cases at 15-19 years of age and five at less than this age. The cell types would be of interest. In Japan there were 56 deaths from cancer of the prostate, 46 of them at 15-29 years of age.
Testis: The peak in testicular cancer among U.S. whites rises sharply from 15-19 years (221 cases) to 25-29 years (801 cases). The total under 30 years of age was 1721. Therapy to date has been very effective, far more so than in Japan where there was an average of 75 deaths a year. If there is indeed a much greater frequency of testicular cancer in the United States than in Japan, the difference may be similar to that between whites and blacks in the United States. Blacks have essentially no peak in the frequency of testicular cancer and are thus apparently resistant to all the main types: seminoma, embryonal carcinoma, and teratoma. Japanese and U.S. whites show a small peak in embryonal carcinoma at 0-4 years, but blacks do not. One wonders why both groups have peaks soon after birth, but only U.S. whites have a second peak in young adulthood.
Ovary: The most interesting feature from the data available was a seemingly high frequency of benign teratomas in Nagasaki Prefecture as indicated by the Tumor Tissue Registry. In 1979, the most recent year for which an annual tabulation had been made, there were 51 cases. Coverage has improved over time since 1972, so the number of cases ascertained has been increasing. The data available show the peak age at diagnosis to be 20-29. Familial cases are thought to be rare. In Willis' series of 50, six were bilateral. One thousand or more should be available for study through the Tissue Registries in Nagasaki and Hiroshima. Through such study, risk factors (personal, familial, or community) and other characteristics of these tumors may be identified.
Tokuoka (Hiroshima) reported that 194 ovarian. cancers had been ascertained in the RERF extended Life-Span Study sample. A statistically significant linear increase was observed with increasing radiation dose but only in 1965-1980, the last half of the study interval. The risk was greatest for females under 20 at the time of the bomb. The minimum latent period was 15-20 years.
A series of 106 benign ovarian tumors ascertained at autopsy also revealed a statistically significant increase in tumors as the radiation dose increased. No particular cell type was associated with radiation exposure with respect either to benign or malignant tumors of the ovary.
Germ cell tumors: The SEER data were readily tabulated. The three histologic types (germinoma, teratoma, and embryonal carcinoma) were about equal in frequency. The gonads were, of course, the most frequently affected organ, with a great male preponderance, 1548 versus 212. The two most frequently affected nongonadal sites also showed a great male preponderance: mediastinum 54 versus 7 and pineal 30 versus 5.
Mori (Tokyo) discussed pineal tumors in Japan. Pineocytoma is rare except in infants. The distribution of 131 cases by cell type was:
|
Number
|
%
|
| Osteosarcoma |
76
|
58
|
| Teratoma, benign |
19
|
14
|
| Teratoma, malignant |
11
|
8
|
| Others |
25
|
19
|
|
131
|
99
|
The peak was at 10-14 years in males, and the sex ratio M/F was 4:1. Koide, in a review of the Autopsy Annuals in Japan, found 3382 brain tumors in a decade, 214 (6.3%) of which were in the pineal. In another series at Tokyo University, Hasagawa found 24 (17.5%) pineal tumors among 137 brain tumors. Other studies indicate that pineal tumors are about 12 times more common in Japan than elsewhere, and the elevation appears to continue after migration to Hawaii. If so, the excess would be attributable to genetic predisposition. In Japan a case-control study with special attention to the family history might be of interest with regard to associated disorders. Mori said he had observed two patients with pineal tumors and maldevelopment of both eyes.
Pierce (Denver) summarized experimental studies in which embryonal carcinoma cells or other cancer cells are injected into blastocysts of the mouse. Trophoectodermal cells in the presence of blastocele fluid regulate the embryonal carcinoma cells. The cancer cells are functionally sperm and, when injected into the ovum, give rise to chimeric mice. This carcinoma has a normal cellular counterpart in the blastocyst, but other cancer cells (e.g., neuroblastoma or melanoma) do not. Neurulation follows blastulation by only a few days, and at this time neuroblastoma cells have been regulated. Also, regulation occurs of melanoma in embryonic skin at the time normal pigment cells arrive in it. He postulated that there may be an embryonic field capable of regulating each carcinoma. If so, an alternate to cytotoxic therapy for cancer may emerge by using naturally occurring specific embryonic inductors to direct differentiation from malignant to benign.
With regard to peculiarities in the occurrence of sex cell cancers in the human, explanations are not available from animal experimentation. We do not know why the rates for choriocarcinoma are so high in Southeast Asia. The neoplasm has not been induced in mice, and the only animal in which it or a similar tumor has been described is the armadillo.
We do not know why whites have a towering peak in the frequency of testicular cancer in adolescents and young adults, whereas nonwhites do not. Pierce said that, in the mouse, cancer of the testis is induced experimentally by a gene.
Can such regulation, as by the blastocyst, explain the low frequency of neuroblastoma in Central Africa, it was asked. Neuroblastoma regresses spontaneously as, for example, the in situ lesions do in virtually all fetuses and as some clinical tumors do, notably type IV-S. Pierce stated that this regression may have an immunologic basis. Little is known about the cell biology. What laboratory studies might be made in Africa None come to mind.
Another puzzle concerns the better prognosis of young versus older children with yolk-sac tumors. Under the age of 4 "embryonal carcinomas" are usually yolk-sac tumors.
Germ cell tumors of all three main types in the SEER Program were about ten times more frequent in males than females at each of the main anatomic sites. Why? Pierce replied that no explanation was available from experimental studies in which the sex ratio cannot be altered, regardless of cell type.
Developmental biology lacks good probes and good ways to "get into" the embryo. The malignant cell and its phenotypic imprint are now being incorporated into normal development, and the fate of the malignant cell can be followed. Pierce said, "I am sure we are going to find in studies of the blastocyst how primary embryonic induction works." It is generally believed that embryonic induction is nonspecific, but by working with known (?) quantities of inductors and taking into account possible changes in receptors compounded by the necessity of cell contact, Pierce thought the problem could be solved.
Breast: The well-known lower rate for breast cancer in Japan than among U.S. women applies, as expected, to women under 30 years of age. Among U.S. white women in this age group, an estimated 679 new cases of breast cancer develop, and among Japanese about 35 deaths occur annually. Notably, 14 women in SEER were 15-19 years of age at diagnosis, and in Japan five deaths from breast cancer were at 10-14 years and three at 15-19 years. Further investigation of those cases would reveal if these tumors were typical breast cancers, and they may be of etiologic interest with regard to the family history, for example. The early deaths in Japan may prove to be due to coding errors with regard to age.
Specimens of breast cancer from the Tumor Tissue Registries in Hiroshima and Nagasaki, along with others from biopsies and autopsies at hospitals in the area, were reviewed by a panel of six experts, three each from the United States and Japan. Tokuoka (Hiroshima) stated that among the 300 specimens, no difference by subtype was found in radiation-induced cancers as compared with those in unexposed women. A strong dose-response relationship was observed among each of the 10-year cohorts exposed at 10-39 years of age, but there was no excess breast cancer among women who were 40-49 years old ATB. Beyond age 50, the dose-response effect was again seen. Recently, when females exposed before 10 years of age reached the usual age for development of breast cancer, they exhibited an excess in frequency that, in time, may exceed that for older age groups ATB. The frequency of bilateral cases seemed to increase with increasing dose.
In response to the question do oral contraceptives increase the frequency of breast cancer, the U.S. side learned that less than 5% of Japanese women use them. They are available only on a physician's prescription, a circumstance which discourages their use.
Sakamoto and Sugano (Tokyo) reported that at the Cancer Institute Hospital in Tokyo, they had studied 63 cases of breast cancer in women 20-29 years of age in 1946-1980. The distribution by subtypes was different than in older groups. The younger women had the highest percentage of papillotubular neo-plasms and the lowest percentage of the scirrhous type. Medullary tubular lesions had a constant rate at all ages. Ten-year survival was less frequent under 30 years of age. Risk factors for breast cancer were not more frequent in this age group, i.e., family history of the neoplasm, bilateralism, or the presence of fibrocystic disease.
Uterine cervix: Because of earlier sexual activity in the United States, the rates of cancer of the uterine cervix are likely to rise earlier in life and be higher there than in Japan. Among women under 30 in the United States, the SEER Program revealed 792 new cases annually. Mortality rates have been declining, however, because of earlier detection through use of the Pap smear and prompt treatment. There were no comparable incidence data for Japan. In the United States, according to Kessler (Baltimore), an increase in the incidence of in situ cancer of the cervix has been accompanied by a decrease in that for invasive carcinoma. To protect against law suits (claiming wrongful diagnosis as benign), there has also been a change in classification, so dysplasias are now frequently called cancer in situ. In addition, exposure has increased to such risk factors as herpes genitalis virus and human papilloma virus.
The U.S. side learned that Pap smears are rarely made in Japan. Women there are reluctant to have pelvic examinations, and self-administered tests are not permitted. Kessler said that serologic test results for the presence or absence of AG-4, an antibody, are correlated with infection by herpes genitalis virus. An increase in AG-4 is associated with first-time infection. The frequency of positive test results is 30% with dysplasia, 95+% with active cancer, and 0% 6 weeks after treatment. Recurrences produce no increase in AG-4.
Females under 10 may develop adenocarcinoma but not squamous cell carcinoma of the cervix. The epidemiology of the two cell types are dissimilar; that for adenocarcinoma is similar to that for cancer of the uterine corpus.
Kessler said that the youngest age at onset of uterine cervical cancer is in the teens among U.S. blacks and the early twenties among whites. The SEER Program, however, ascertained four cases annually among whites 10-14 years of age and 18 cases at 15-19 years. Marriage to a man whose previous wife had this cancer increases the risk 2.5-fold for the new wife.
Present knowledge about viruses in the etiology of cervical cancer indicates that either herpes type I or II may be involved. The evidence for a role of human papilloma virus is still weak. Types 6 and 11 have been derived from cervical cancer.
Vagina: The SEER Program listed 48 women with vaginal carcinoma. The years covered, 1973-82, followed the discovery in 1971 that maternal use of diethylstilbestrol (DES) during pregnancy caused about 1 in 1000 daughters to develop clear cell adenocarcinoma of the vagina or cervix at 8-29 years. The 42 cases at 15-29 years of age registered by the SEER Program may well be of this type, which was previously very rare so early in life. This suggests that early onset of other adult-type tumors may be due to transplacental carcinogenesis.
Vulva: The SEER Program listed 32 women under 30 years with cancer of the vulva. All of them were 15-29 years of age. A case control study might reveal risk factors more easily than studies of larger samples of affected older women, whose longer life histories may make etiologic influences less discernible.
Choriocarcinoma: In the SEER Program carcinoma of the placenta was observed in 78 white women under age 30. Except for one diagnosed at 10-14 years, all occurred at 15-29 years. The high frequency of choriocarcinoma among women in China and neighboring countries is well known. A high proportion are cured by chemotherapy, even when metastases are extensive. In Japan 386 deaths occurred, 1968-1978, among women under 30 with this cancer. Kajii (Ube) stated that the number of deaths in Japan from choriocarcinoma has declined from 320 in 1960 to 56 in 1981. The frequency of hydatidiform moles, a precursor, is 1 in 1200 births in the United States, 1 in 300 in Japan, and 1 in 100 in Southeast Asia. The Trophoblastic Disease Registry in Japan in 1974-1979 listed 10,989 hydatidiform moles, 516 invasive moles, and 330 choriocarcinomas. Moles increase in frequency with advancing maternal age. The effect of migration to the United States was not clear with respect to the frequency of hydatidiform moles among Japanese or people from Southeast Asia.
Because most complete moles arise from duplication of a haploid sperm in a functionally empty ovum, malignant transformation is thought to result from homozygosity of a recessive mutant gene. Normally, the gene controls cell growth, but upon mutation it can no longer do so, and the malignant growth may occur (Kajii, T., et al., Amer. J. Obs. Gyn. 150:57-64, 1984).
Urinary bladder: Cancer of the urinary bladder affects 134 U.S. white males and 60 white females under age 30 annually according to SEER data. All but nine were 10 years of age or older. The cell type was not available at the time of the meeting, but case series previously reported indicate that virtually all were transitional cell carcinoma. Cohen (Omaha) stated that these studies had shown no clustering of cases, excess of familial occurrence, or relation to known (or suspected) carcinogens that affect the bladder (cigarette smoking, occupational exposure, analgesic abuse, or cyclophosphamide therapy).
Cohen stated that the histology is the same in both countries. Transitional cell carcinoma may be (1) papillary, 75-90% of all bladder cancer, which is usually noninvasive and recurs frequently—50% of patients have a recurrence in 1 year and 90% in 5 years, or (2) nonpapillary, which has bizarre cells in high-grade lesions--treatable but often invasive before diagnosis, so the 5-year survival rate is low.
Bladder carcinoma under age 30, according to previous studies, is almost limited to males, but the SEER data shown above indicate that in U.S. whites the sex ratio, 2.5:1, is similar to that observed in older age groups. Experimental studies show that in utero exposure to BBN or BVN causes bladder cancer, the cell type of which, papillary or not, depends on the dose. Promoters, such as saccharin or caffeine, prenatally should have little effect because the cell proliferation they can cause is likely to have no effect on cells that are already proliferating in the fetus at maximal speed. Instead, one should look for prenatal influences among exposures, such as cigarette smoke, which interact with DNA. Low-for-date birth weight can be used as an indication of the effect of maternal use of cigarettes during pregnancy in connection with transitional cell carcinoma of the bladder in young adults. Again, the SEER cases might, through a case control study, be revealing in this regard.
With regard to therapy, isolated papillary lesions can be cured by fulguration. One cannot tell whether an initial benign papillary lesion is an isolated occurrence, so a means for distinguishing nonrecurrent benign tumors from those destined to recur as malignant neoplasms is much to be desired.
Hematopoietic System
Differences between the United States and Japan in the frequencies of cancers of the blood-forming organs have been evaluated by several recent workshops under the binational Cooperative Cancer Research Program. The focus has been on a reciprocal relationship between certain lymphoreticular neoplasms (low frequency in Japan) and certain autoimmune diseases (high frequency in Japan). The neoplasms are nodular lymphoma; Hodgkin's disease under 30; acute lymphocytic leukemia in children, which did not develop a peak in frequency at 4 years of age until 20 years after the peak developed in the United States; and chronic lymphocytic leukemia in adults. The autoimmune diseases known to be more frequent in Japan than in the United States include systemic lupus erythematosus (SLE), Hashimoto's thyroiditis, and Takayasu's aortitis. Presumably, the function or proportion of OKT4 to OKT8 lymphocytes is different in Japanese than in U.S. whites. In effect, the Japanese have greater protection against the lymphomas and increased susceptibility to the autoimmune diseases. A similar excess of SLE has been found in Hawaii among all Asians. Namba (Hiroshima) described how he used the Tumor Tissue Registry to detect the low frequency of nodular lymphoma. Similar review by Namba of specimens on file in Honolulu was encouraged, comparing whites with the several Asian groups there. The histologic study, as Sugano put it, would then be made "by the same eyes."
Namba concluded from his studies that in Hiroshima, under 30 years of age, there were high percentages of Hodgkin's disease and lymphoblastic lymphoma, as well as extranodal origin of diffuse lymphoma, immunoblastic lymphoma, and small noncleaved lymphomas. These findings suggested an excess of primitive T cell lymphomas in Japan and that nodular lymphomas are relatively uncommon not only in adults but also in children there.
The SEER data revealed six cases of multiple myeloma under age 30. The youngest was a boy 9 years of age, whose diagnosis was confirmed by electrophoretic studies and bone marrow study. In Japan 56 deaths under 30, 1968-1978, were ascribed to multiple myeloma, the diagnosis of which is notoriously variable in its accuracy.
The SEER data also detected 12 cases of Kaposi's sarcoma, nine of them in single males living in San Francisco. The diagnoses were made in 1981-1982, and illustrate that luck is needed in establishing registries for detecting geographically localized clusters. It also exemplifies how a small excess of a rare cancer in a registry can be etiologically important.
Thyroid Cancer
The SEER data showed that thyroid cancer was the diagnosis for 300 males and 1275 females under 30. The sex ratio is thus 1:4.25, similar to that in Japan. A. Sakamoto (Tokyo) had studied cases from the Thyroid Cancer Registry in Japan, which had accessioned 6012 cases, 1977-1982, 736 of which were under 30 years of age. This resource, previously unknown to the U.S. participants, may afford the opportunity to use record linkage to find persons in the fixed study samples of RERF who developed thyroid cancer after they migrated from Hiroshima or Nagasaki.
In another series study by Sakamoto, among 56 cases under age 30 seen at the Cancer Institute in Tokyo, 10 (18%) had chronic thyroiditis. This seems to be an important subset from which further understanding of the relationship between the two diseases can be explored.
The Japanese have a great excess of latent carcinomas of the thyroid incidentally discovered at autopsy. There is no age gradient. Stemmermann pointed out that some of the latent lesions were partially or totally sclerosed, suggesting that they arose (and were frozen in place) long ago.
Children's Cancer
Kobayashi (Tokyo) presented data from the All-Japan Children's Cancer Registry which covers six metropolitan areas, 1969-1982. These data have long indicated a deficiency in the frequency of Wilms' tumor (WT) as measured against that for neuroblastoma (Nb). In the SEER Program the ratio Nb/WT was 441/319 - 1.38. In Japan it was 1691/800 =2.11. Wilms' tumor has a relatively constant occurrence throughout the world, although an excess among blacks in the Philadelphia area has recently been reported. The question, then, is whether in Japan the rate of Nb is high or that for WT is low. The rates are not available as yet.
Retinoblastoma (Rb) appears to be much more common in Japan than in the United States. The relative frequencies of Nb, Rb, and WT in each country are:
|
SEER
|
JAPAN
|
| Nb |
49.6
|
46.5
|
| Rb |
14.5
|
31.5
|
| WT |
35.9
|
22.0
|
|
100.0
|
100.0
|
Rb varies markedly in frequency geographically. The rates are very high in India and Pakistan. Nb varies in the other direction; it is virtually absent in Central Africa. It appears that there is an excess of Rb and a deficiency of WT in Japan.
Kobayashi said (personal communication) that congenital aniridia with Wilus' tumor occurred only half as often in the Japanese series as in western countries. Among the 800 patients with WT, only four or five instead of 10 had aniridia, which is due to partial deletion of the short arm of chromosome 11. Retinoblastoma occurs in a syndrome due to partial depletion of the long arm of chromosome 13.
Variation in rates for these cancers may be due to differences in an inherent capacity to maintain chromosomal integrity. Such differences have been reported by Mitelman with regard to chromosomes associated with certain other cancers.
Latent Periods
Land (Bethesda) spoke of latent periods after exposure to ionizing radiation. The atomic bomb exposure in Japan and exposure to Ra-224 (half-life = 4 days) for therapy of bone tuberculosis in West Germany involved short-term exposures followed by a wave of cancer--leukemia and osteosarcoma, respectively. Other cancers did not have a wave-like pattern, e.g., cancer of the breast, lung, or thyroid. Cohen pointed out that cells have to be dividing, or cancer should not be inducible. Retinal cells stop dividing early in life so retinoblastoma would not be expected as a radiation effect. Neurons and cardiac muscle cells also stop dividing early in life. Why, then, do cells that give rise to benign prostatic hypertrophy not become malignant when exposed to ionizing radiation? And why with the life-long rapid turnover of cells in the small intestine do they rarely become malignant? The response: there is more to carcinogenesis than the rapid turnover of cells.
In Perspective
The workshop generated many new ideas for study concerning binational differences in type-specific cancers, as well as clusters within each country. The suggestions, as summarized in this report, are too numerous to list here.
In the past SEER data were used to compare the age distribution of epithelial and nonepithelial cancers. In childhood, nonepithelial cancers account for about 90% of all malignancies. At 15 years of age a sharp rise in rates begins and lifts the frequency of epithelial cancers to equal that for non-epithelial cancers at 20 years of age. The rise continues so that at 30 years of age the ratio of epithelial to nonepithelial cancers is 7:1, a difference which is maintained for the rest of life as both rates climb in parallel.
Cancers in adolescence are a mixture of nonepithelial cancers that are declining from peaks in childhood, of carcinomas that are beginning to rise to peaks later in life, and of neoplasms that are special to adolescence and peak then (e.g., bone sarcoma).
Case control studies of cancers in these three ages of youth in comparison with one another and with studies of older persons should be informative with regard to etiology and pathogenesis.
SEMlNAR AGENDA AND PARTICIPANTS
(1) US-JAPAN WORKSHOP ON THE ETIOLOGIC IMPORTANCE OF CANCER SUBTYPES
Honolulu, Hawaii, October 16-17, 1984
AGENDA
| Tuesday, October 16 |
| 1. |
Clues to Cancer Etiology from Rare Subtypes |
Frederick P. Li, M.D.
Head, Clinical Studies Section
CEB, NCI, Bethesda, MD |
| 2. |
Subtypes in Relation to Maps of Cancer Genes |
John J. Mulvihill, M.D.
Head, Clinical Genetics Section
CEB, NCI, Bethesda, M.D. |
|
COFFEE BREAK |
|
| 3. |
Tumor Cell Atypism and Automated Cytology |
Noboru Tanaka, M.D.
Director, Chiba Cancer Center Research Institute |
| 4. |
Basic versus Variable Subtypes |
Haruo Sugano, M.D.
Director, Cancer Institute
Tokyo, Japan |
|
LUNCH |
|
| 5. |
Epidemiologic Implications of Cancer Histology and Subsite Differences Between Caucasians and Japanese |
John W. Berg, M.D.
Professor of Pathology
University of Colorado |
| 6. |
Some Subtype Differences between Caucasians and Japanese in Hawaii |
Grant N. Stemmermann, M.D.
Professor of Pathology
University of Hawaii |
|
COFFEE BREAK |
|
| 7. |
Subtypes of Breast Cancer in Japanese as Compared with Caucasians |
Goi Sakamoto, M.D.
Associate Member
Department of Pathology
Sapporo Medical College |
| 8. |
Subtypes of Lung Cancer and their Morphogenesis |
Ynkio Shimasato, M.D.
Chief, Department of Pathology
National Cancer Center
Tokyo, Japan |
Wednesday, October 17 |
| 9. |
Subsets of Lymphomas and Leukemia |
J.N.P. Davies, M.D.
Medico-Legal Consultant
Albany, New York |
| 10. |
Monoclonal Antibodies and Subtypes of Lymphoma |
Kokichi Kokushi, M.D.
Professor of Pathology
Sapporo Medical College |
|
COFFEE BREAK |
|
| 11. |
Subtypes of Stomach Cancer and Their Epidemiology |
Aya Hanai, B.S.
Section Head
The Center for Adult Disease
Osaka, Japan |
| 12. |
Subtypes of Rhabdomyosarcoma |
William A. Newton, Jr., M.D..
Professor of Pathology
Columbus Children’s Hospital
Ohio |
|
LUNCH |
|
| 13. |
The Importance of Hormones as Tumor Markers |
Hiroo Imura, M.D.
Professor of Medicine
Kyoto University |
| 14. |
Hormone Receptors and Subtypes of Endometrial Cancer |
Junzo Kato, M.D.
Professor of Gynecology
Yamanashi Medical University |
| 15. |
General Discussion: Suggestions for Future Research |
Robert W. Miller, M.D. and Haruo Sugano, M.D. |
PARTICIPANTS
UNITED STATES
John W. Berg, M.D.
Professor of Pathology
Campus Box B-216
Colorado Medical Center
4200 East 9th Avenue
Denver, CO 80262
(303) 394-7636
J. N. P. Davies, M.D.
Medical-Legal Consultant PC
5 Pine Street
Albany, NY 12207
(518) 463-2136
Frederick P. Li, M.D.
Dana-Farber Cancer Center
44 Binney Street, Room 1110
Boston, MA 02115
(617) 732-3158
Robert W. Miller, M.D.
Chief, Clinical Epidemiology Branch
8C41 Landow Building
National Cancer Institute, NIH
Bethesda, MD 20892
(301) 496-5785
John J. Mulvihill, M.D.
Chief, Clinical Genetics Section
8C41 Landow Building
National Cancer Institute, NIH
Bethesda, MD 20892
(301) 496-4947
William A. Newton, M.D.
Children's Hospital
700 Children's Drive
Columbus, OH 43205
(614) 461-2310
Grant N. Stemmermann, M.D.
Clinical Professor of Pathology
University of Hawaii Medical School
Kuakini Hospital
Honolulu, HI 96817
(808) 235-1693
JAPAN
Aya Hanai, B.S.
Section Head
The Center for Adult Disease, Osaka
Nakamichi, Higashinari-ku
Osaka 537
(06) 972-1181
Hiroo Imura, M.D.
Professor
Department of Medicine
Kyoto University
Yoshida, Sakyo-ku
Kyoto 606
(075) 751-3111
Junzo Kato, M.D.
Professor
Department of Gynecology
Yamanashi Medical University
Tamaho-mura, Nakakoma-gun
Yamanashi 409-38
(0552) 73-1111
Kokichi Kikuchi, M.D.
Professor
Department of Pathology
Sapporo Medical College
Chuo-ku
Sapporo 060
(011) 611-2111
Goi Sakamoto, M.D.
Associate Member
Department of Pathology
Kami-Ikebukuro, Toshima-ku
Tokyo 170
(03) 918-0111
Yukio Shimosato, M.D.
Chief
Department of Pathology
National Cancer Center
Tukiji, Chuo-ku
Tokyo 104
(03) 542-2511
Haruo Sugano, M.D.
Director
Cancer Institute
Kami-Ikebukuro, Toshima-ku
Tokyo 170
(03) 918-0111
Noboru Tanaka, M.D.
Director
Chiba Cancer Center Research Institute
Nitona-cho
Chiba 280
(0472) 64-5431
(2) US-JAPAN COOPERATIVE CANCER RESEARCH PRODGRAM WORKSHOP ON ADULT-TYPE CANER UNDER 30 YEARS
Tokyo, Japan, March 11-13, 1985
AGENDA
| March 11, 1985 |
| 8:45 |
WELCOME |
Dr. Sugano and Miller |
| 9:00 |
Epidemiologic Perspective of the Japanese Experience |
Dr. Kunio Aoki |
| 9:40 |
U.S. Morbidity: SEER Program |
Dr. John L. Young, Jr. |
| 10:20 |
COFFEE BREAK |
|
| 10:40 |
Morbidity in Hiroshima: The Tumor Tissue Registry |
Dr. Scott Davis |
| 11:20 |
The All-Japan Autopsy Registry |
Dr. Yoshinori Urano |
| 12:00 |
LUNCH |
|
| 2:00 |
The All-Japan Childhood Cancer Registry |
Dr. Noboru Kobayashi |
| 2:40 |
Xeroderma Pigmentosum and Skin Tumors |
Dr. Hiraku Takebe |
| 3:20 |
COFFEE BREAK |
|
| 3:40 |
Bone and Soft Tissue Tumors |
Dr. Kotaro Furuya |
| 4:20 |
Osteosarcoma versus Ewing's Tumor: SEER Data |
Dr. Robert W. Miller |
| 5:00 |
ADJOURN |
|
March 12, 1985 |
| 9:00 |
Pineal Tumors in Japan |
Dr. Wataru Mori |
| 9:40 |
Sex Cell Tumors |
Dr. G. Barry Pierce |
| 10:20 |
COFFEE BREAK |
|
| 10:40 |
Hydatidiform Mole and Choriocarcinoma |
Dr. Tadashi Kajii |
| 11:20 |
Pigmented Tumors |
Dr. Tsutomu Kasuga |
| 12:00 |
LUNCH |
|
| 2:00 |
Uterine Cervix |
Dr. Irving I. Kessler |
| 2:40 |
Stomach Cancer |
Dr. Pelayo Correa |
| 3:20 |
COFFEE BREAK |
|
| 3:40 |
Colon Cancer |
Dr. Grant N. Stemmermann |
| 4:20 |
Adenomatosis and Polyps of the Colon |
Dr. Joji Utsunomiya |
| 5:00 |
ADJOURN |
|
March 13, 1985 |
| 9:00 |
Breast Cancer |
Dr. Haruo Sugano |
| 9:40 |
Breast Cancer and Ovarian Cancer |
Dr. Shoji Tokuoka |
| 10:20 |
COFFEE BREAK |
|
| 10:40 |
Cancer of the Urinary Bladder |
Dr. Samuel M. Cohen |
| 11:20 |
Lymphoma Subtypes: Japarnversus U.S. |
Dr. Koji Nanba |
| 12:00 |
Thyroid Cancer |
Dr. Atsuhiko Sakamoto |
| 12:40 |
LUNCH |
|
| 2:00 |
Some Thoughts for the Future Biostatistics Medicine |
Dr. Charles E. Land
Dr. Andrey K. Brown |
| 3:00 |
General Discussion |
Drs. Miller and Sugano |
| 4:00 |
ADJOURN |
|
PARTICIPANTS
UNITED STATES
Robert W. Miller, M.D., Co-chair
Clinical Epidemiology Branch
8C41 Landow Building
National Cancer Institute, NIH
Bethesda, MD 20892
(301) 496-5785
Audrey K. Brown, M.D.
Professor of Pediatrics
Downstate Medical Center
450 Clarkson Avenue, Box 49
Brooklyn, NY 11203
(718) 270-1692
Samuel M. Cohen, M.D.
Dept. of Pathology and Lab. Medicine
Univ. of Nebraska School of Medicine
42nd and Dewey Avenue
Omaha, NE 68105
(402) 559-4186
Pelayo Correa, M.D.
Department of Pathology
LSU Medical Center
1901 Perdido Street
New Orleans, LA 70112
(504) 568-6035
Rufus S. Day, Ph.D.
Acting Chief, Protein Section
Lab. of Molecular Carcinogenesis
Building 37, Room 3C23
National Cancer Institute, NIH
Bethesda, MD 20892
(301) 496-1470
Irving I. Kessler, M.D.
Chairman, Department of Social and Preventive Medicine
Univ. of Maryland School of Medicine
655 West Baltimore Street
Baltimore, MD 21201
(301) 528-7866
Charles E. Land, Ph.D.
Radiation Epidemiology Branch
3A22 Landow Building
National Cancer Institute, NIH
Bethesda, MD 20892
(301) 496-6600
G. Barry Pierce, M.D.
University of Colorado Medical Center
4200 E. 9th Avenue
Denver, CO 80262
(303) 394-7580
Grant N. Stemmermann, M.D.
Clinical Professor of Pathology
Univ. of Hawaii Medical School
Kuakini Hospital
Honolulu, HI 96817
(808) 235-1693
John L. Young, Jr., Ph.D.
Demographic Analysis
Biometry Branch, NCI, NIH
Blair Building, 532-B
Bethesda, MD 20892
(301) 427-8829
Scott Davis, Ph.D.
Radiation Effects Research Foundation
5-2 Hijiyama Park
Hiroshina 730, JAPAN
K. Mabuchi, M.D., Ph.D.
Departiaent of Epidemiology
Radiation Effects Research Foundation
5-2 Hijiyama Park
Hiroshima 730, JAPAN
JAPAN
Kunio Aoki, M.D.
Professor
Department of Preventive Medicine
Nagoya University
School of Medicine
65 Tsurumaicho, Showa-ku
Natoya 466
(052) 741-2111
Kohtaro Furuya, M.D. '
Professor
Department of Orthopedic Surgery
Tokyo Medical and Dental University
1-5-45, Yushima, Bunkyo-ku
Tokyo 113
(03) 813-6111
Takayoshi Ikeda, M.D.
Professor
Department of Pathology
Nagasaki University
School of Medicine
12-4 Sakamoto-cho
Nagasaki 852
(0958) 47-2111
Tadashi Kajii, M.D.
Professor
Department of Pediatrics
Kogushi, Nishi-ku
Ube 755
(0836) 22-2111
Tsutomu Kasuga, M.D.
Professor
Department of Pathology
Tokyo Medical and Dental University
Faculty of Medicine
1-5-45, Yushima, Bunkyo-ku
Tokyo 113
(03) 813-6111
Noboru Kobayashi, M.D.
General Director
National Children's Medical Research Center
3-35-31, Taishido
Setagaya-ku, Tokyo 154
(03) 414-8121
Wataru Mori, M.D.
Professor
Department of Pathology
Faculty of Medicine
University of Tokyo
7-3-1, Hongo, Bunkyo-ku
Tokyo 113
(03) 812-2111
Koji Nanba, M.D.
Professor
Faculty of Integrated Arts and Science
Hiroshima University
1-1-89, Higashi-senda-machi
Naka-ku, Hiroshima 730
(082) 241-1221
Atsuhiko Sakamoto, M.D.
Associate
Department of Pathology
Cancer Institute
1-27-1, Kami-Ikebukuro
Toshima-ku, Tokyo 170
(03) 918-0111
Haruo Sugano, M.D.
Director
Cancer Institute
1-37-1, Kami-Ikebukuro
Toshima-ku, Tokyo 170
(03) 918-0111
Hiraku Takebe, Ph.D.
Professor
Dept. of Experimental Radiology
Faculty of Medicine
Kyoto University
Yoshida-konoecho, Sakyo-ku
Kyoto 606
(075) 75-2111
Shoji Tokuoka, M.D.
Professor
Department of Pathology
Hiroshima Univ. School of Medicine
1-2-3, Kasumi, Minami-ku
Hiroshima 734
(082) 251-1111
Yoshinori Urano, M.D.
Professor
Department of Pathology
Faculty of Medicine
University of Tokyo
7-3-1, Hongo, Bunkyo-ku
Tokyo 113
(03) 812-2111
Joji Utsunomiya, M.D.
Professor
Department of Surgery
Hyogo College of Medicine
1-1, Mukogawa-cho
Nishinomiya 663
(0798) 45-6111
