REPORTS ON SEMINARS

(1) Workshop on “Genetic Disorders with High Risk of Cancer”

A workshop on the different arrays of cancers in genetic syndromes was convened at the Hawaii Prince Hotel in Honolulu on February 6-7, 1995, by Drs. Robert W. Miller and Haruo Sugano. There were 6 U.S. and 7 Japanese participants. The purpose was to consider the possible mechanisms to account for the marked differences in the diverse cancers in each syndrome.
Dr. Makoto Goto described his study with Drs. Miller, Sugano, and Ishikawa of 123 case-reports in Japan on Werner’s (adult progeria) syndrome (WS) and neoplasia: 24 soft-tissue sarcomas, 8 osteosarcoma, 21 melanoma, 15 meningioma, 12 leukemia (most specified as AML) or preleukemia, and 23 thyroid cancer predominated. The ratio of non-epithelial cancer to epithelial cancer, instead of the usual 10:1 was 1:1.5. Five of the melanoma were intranasal, two of them in sisters, and most of the rest were acral, on the soles of the feet, a very unusual distribution by cell type. Other neoplasia that clustered in first-degree relatives included three siblings with soft tissue sarcoma of the elbow, and two brothers with soft-tissue tumors of the leg. The large number of cases known in Japan is attributable to inbreeding revealing an autosomal recessive trait and ease of publishing case-reports.
Dr. Warren W. Nichols noted that a gene for WS, mapped to chromosome 8p, is near a DNA repair gene (DNA polymerase!!!), but recent studies have shown that it is not the WS gene. Haplotype gene-markers for the WS gene differ in Japanese and Caucasians, which may account for the lack of an increase in melanoma and thyroid cancer in Caucasians with WS. Intragenic mutations may account for the concordance of cancer sites and types in first-degree relatives. It has been proposed that the gene counts cell cycles and in WS stops at 20 instead of 60. In any event, a defective mechanism for regulating cell cycle may be responsible for the odd array of cancers in WS.
Dr. James German described 197 cancers in 165 persons with Bloom’s syndrome (BS) (dwarfism, sun-sensitive facial rash and chromosomal instability). In childhood, leukemia and lymphoma were excessive. Five children had MDS. In adults the frequency of carcinoma was increased, especially of the gastrointestinal tract (12 colon cancers), including four with carcinoma at the base of the tongue. Seven had breast cancer. Sixty-five had single cancers, 17 had multiple, including several with triple primaries of the gastrointestinal tract. The oldest patients with the syndrome are only 46, so the time of carcinoma occurrence is much advanced compared with normal. The increased frequency of sister-chromatid exchanges (SCEs), typical of the syndrome, presumably is related to oncogene activation. The genomic instability allows mutations to arise spontaneously.
Dr. Hiraku Takebe described the first 11 cases of BS in Japan, validated by demonstrating an excess of SCEs, and examination by Dr. German. The sun-sensitive facial rash seen in occidental cases does not appear clearly in Japanese. In 4 of 9 families the parents were consanguineous. BS, not validated cytologically, has been reported in China and two validated cases are known in Tunisia.
Dr. Kouichi Tatsumi added information on the genetic Instability in BS cells. Recently his group has performed fluctuation tests using a flow-cytometric mutation assay. It showed spontaneous loss of an HLA-A allele expression in T-lymphocytes from a BS patient that was about ten times higher (5 x 10^-4/cell/division) than those in T-1ymphocytes from normal controls. These and other observations indicate that the mutator phenotype of BS cells is mainly due to an increased frequency of large DNA alterations, namely deletions and recombinations, reflecting the profound genomic instability to which cancer proneness in BS may be attributed.
Dr. Blanche P. Alter described Fanconi’s anemia (FA) as a premalignant disorder. Five complementation groups have been identified by cell fusion studies. Among about 1000 cases reported in the literature, 156 had one or more malignancies, for a total of 171. The relation of complementation group to cancer is not known; l0-15% have type C. There were 82 leukemia, 31 myelodysplasia, 37 liver neoplasms, and 46 other cancer. All leukemias were myeloid except for two lymphoid (L-1). They developed on average at 15 years of age; 25% presented without previous aplastic anemia. Only about 5% survived the leukemia (when last reported). Five percent of patients with FA developed other tumors after marrow transplantation -- a higher frequency than in other marrow transplant recipients. Liver tumors were reported in 37 patients (29 had cancer), at an average age of 16 years. All had had androgen therapy for their aplastic anemia, indicating that the liver tumors were due to the hormone treatment in this predisposing genetic disease. The tumors were generally not invasive or metastatic. No deaths were due to liver tumor. Liver tumors had as second primaries 5 leukemia and two other cancer. The other cancers reported were 17 oropharyngeal (8 tongue!), 16 gastrointestinal (8 esophageal), 11 gynecological, 4 brain and 8 others. In about 50 cases of other cancers, 6 had multiple primaries. Approximately 20 were oropharyngeal, 15 were gastrointestinal, and 10 were gynecological. The term “preleukemia” has been applied to FA with myelodysplastic bone marrow and clonal cytogenetic abnormalities. Systematic bone marrow examinations of all FA patients have found about 30% each with myelodysplasia and cytogenetic markers, with no correlation between the two. Some of the abnormal clones have been transient and others have persisted for up to 13 years. FA treated with bone marrow transplants may be at increased risk of second cancers following immunosuppression and irradiation (3 cases to date in 100+ transplant recipients). The mechanism for carcinogenesis is presumably chromosomal instability which may lead to loss or inactivation of genes such as tumor suppressors.
Dr. Masao S. Sasaki noted in his presentation that the variable clinical manifestations makes FA difficult to diagnose. Erroneous diagnoses not associated with cancer may diminish the estimate of the cancer risk in true FA. 123 cases of FA in Japanese have been reported in the literature, 12 had cancer, 5 preleukemia or MDS, 3 ANLL, 2 squamous carcinoma (sibs), I hepatocellular carcinoma and 1 Wilms’ tumor. Recently the FA gene (FACC gene) responsible for the complementation C group has been cloned. Its function is unknown, but it makes possible the molecular identification of one form of FA The future holds the promise for characterization of the genetic heterogeneity of FA and associated biological conditions, including neoplasia.
The Japanese Registry of FA patients may not be known outside Japan. Dr. Alter suggested that It become a member of the International FA Registry, and after the meeting she communicated this thought to Dr. Yukiko Tsunematsu of the National Children’s Hospital, who now plans to assist in making this link.
Dr. Akira Horii and his associates reasoned that there may be human homologues of yeast and bacterial genes disrupted in mismatch repair (e.g., MSH2, MLH1, PMS1 and PMS2), with significant increases in the incidence of replication errors (RERS) of microsatellite markers. Individuals with mutations in one of these homologues may have an increased likelihood of accumulating genetic alterations that lead to transformation of normal cells to cancer cells and the occurrence of multiple primary cancers, as in hereditary non-polyposis cancer of the colon (HPNCC). Horii et al isolated 6 human genes homologous to but distinct from hPMS1 and hPMS2, termed hPMS3 through hPMS8, some of which have been mapped to chromosome bands 7q11.23 and 7q22. Mutations of these genes are being sought in patients with HPNCC or the types of cancer associated with it. To investigate whether genetic defects involving the mismatched repair system are an important risk factor in patients with multiple primary cancers, RERs were examined at microsatellite loci in 79 primary cancers in 38 patients. The RER(+) phenotype was found at 5 microsatellite loci in tumors from 34 of 38 (89%) of patients with multiple primary cancers, but in only 19 of 174 (11%) patients with single primary cancers. Testing for RER may be a means for detecting patients at high risk of second primary cancers.
Dr. Miller reported on the neoplasms found in Rubinstein-Taybi syndrome (RTS) of broad thumbs and toes and characteristic facial anomalies. Of 724 cases of RTS entered in a registry maintained at Cincinnati Children’s Hospital, 36 are known to have tumors. Eleven were in the nervous system, including 2 each of oligodendroglioma, medulloblastoma, neuroblastoma and meningioma and 1 each of pheochromocytoma, neurilemmoma, ectopic pinealoma, and pituitary growth hormone cell microadenomas. Other tumors of the head or neck were 2 rhabdomyosarcoma of the nasopharynx, and one each of composite odontoma of the mandible, angiofibroma of the eyelid, choristoma of the eyelid and orbit, dermoid cyst of the eye, preauricular pilomatrixoma, multiple pilomatrixoma and epidermoid cyst of the parotid. It is of interest that these neoplasms are of tissue that is derived from the neural crest, or arise from developmental errors, most frequently in the head, which is severely malformed in the syndrome. The pathogenesis may be similar to that for neoplasia in neurofibromatosis type 1 except for its odd excess of leukemia (see Shannon, below), plus tumors arising from developmental errors. About 5% of patients with NF1 develop cancer. The percentage is lower in RTS, but the array of tumors and their relative frequencies differ from normal. In other cancer-prone genetic diseases the frequencies are about 25% in Werner’s syndrome at 30-60 years of age (see Goto, above), 36% in Bloom’s syndrome (see German, above), at least 15% in Fanconi’s anemia (Alter, above), and 50% in persons under 45 years of age with Li-Fraumeni syndrome. In each disorder the array of cancers is different, due no doubt, to dissimilar carcinogenic mechanisms.
Dr. Kevin M. Shannon spoke on characterizing the genetic and biochemical basis of heritable predisposition to myelodysplastic syndrome (MDS) and myelogenous leukemia. Among patients with juvenile CML (JCML) and/or monosomy 7, ten percent have neurofibromatosis type 1 (NF1). Also associated with an increased risk of AML and MDS are Fanconi anemia, Schwachmann-Diamond syndrome, and certain other heritable factors. His group has demonstrated that the NF1 gene (NF1) encodes the protein, neurofibromin, which accelerates the hydrolysis of GTP on p21^ras. RAS mutations are common in myeloid leukemia, an observation that suggests NF1 negatively regulates hematopoietic cell growth through its effect on p21^ras. Loss of normal NF1 was found in bone marrows of about half of the children with NF1 and myeloid leukemia. His group demonstrated activating RAS mutations in 12 of 55 children with preleukemia but in none of 19 patients with NF1. These results are consistent with the action of NF1 as a tumor-suppressor gene in myeloid cells. Very few children with NF1 (1% or less, and boys more than girls) develop leukemia. Some with leukemia show monosomy 7 as well as loss of the normal NF1 allele. The same disorder is not seen in adults, indicating that multiple genetic events are required for leukemogenesis in NF1, and that NF1 is required only to regulate hematopoietic cell growth in early life (developmentally restricted tumor-suppressor activity). Other studies have shown that children with severe congenital neutropenia are at risk of developing MDS and AML while receiving recombinant granulocytic strangulating factor, and that bone marrows of children with an inherent predisposition to myeloid leukemia acquire the same secondary genetic alterations as in sporadic cases. A review of the role of childhood monosomy 7 (epidemiology, biology and mechanistic implications) was published soon after the workshop (Luna-Friedman S, Shannon KM, Lange BJ: Blood April 1995), which concluded with a model for the role of monosomy 7 in the development of malignant myeloid disorders.
With regard to the difference in the sex ratio pre and postconception, as shown in the diagram, Dr. Goto looked at his data on Werner’s syndrome with AML and preleukemics (12 Japanese cases). The sex of 11 was known. A11 were males. Two of four Caucasians with both disorders, however, were females.
Dr. Takehiko Sasazuki and his group have been studying germ-1ine and somatic mutations in colon cancers from cancer-prone families. They compared heritable non-polyposis colon cancer with sporadic non-polyposis colon cancer under age 45 and found no mutations in the hMLH1 gene. They found one patient in whom genetic instability was due to complete loss of functional hMSH2 by a germline and a somatic mutation, which may, therefore, be involved in carcinogenesis in cancer-prone families and in some sporadic tumors. Somatic mutations of the Ki-ras gene were found in 25% and 19% of tumors with and without genomic stability, respectively. In contrast, somatic mutations in the p53 and APC genes were found in 25% and 0% in genomic instability-positive tumors, less frequently than in genomic instability negative tumors (43% and 33%). These observations suggest that tumor suppressor genes other than p53 and APC may also be preferentially affected in tumors with genomic instability.
Dr. William G. Nelson spoke on p53 and the cellular response to DNA damage. All cancer-prone syndromes may be related to the cell’s response to stress. Genotoxic insults, such as exposure to ionizing radiation, cause cell injury. The cell signals its injury, and the response may be quiescence, repair, proliferation, cell death, transformation to cancer or differentiation. One particular cell signal involves p53, which responds to chromosomal strand-breaks. It is the most commonly mutated gene and a major determinant of the cell’s fate after DNA damage. The p53 protein turns on GADD45 (Growth-arrest-and-DNA Damage-inducible), whose protein stimulates excision repair. Meanwhile, p53 has blocked cell cycle progression and triggered programmed cell death. It blocks the cell cycle presumably by stimulating activity of the p21 gene, whose protein inhibits cyclin-dependent protein kinases, which are key enzymes for cell progression.

PARTICIPANTS

UNITED STATES

Dr. Blanche P Alter
Division of Pediatric Hematology/Oncology
University of Texas Medical Branch
Galveston, TX 77555
Tel: (409) 772-2341
Fax: (409) 772-4599

Dr. James L. German
New York Blood Center
310 E. 67 Street
New York NY 10021-6204
Tel: (212) 570-3075
Fax: (212) 570-3195

Dr. Robert W Miller
National Cancer Institute
EPN-400
Bethesda MD 20892-7360
Tel: (301)496-5785
Fax: (301) 496-1854

Dr. William G Nelson
Marburg 411
Johns Hopkins Hospital
600 N Wolfe Street
Baltimore, 21287
Tel: (410) 614-1663
Fax: (410) 955-0833

Dr. Warren W. Nichols
Merck, Sharpe and Dohme Research Labs
Bldg 44-1
West Point, PA 19486
Tel: (215) 656-7980
Fax: (215) 652-3888

Dr. Kevin M. Shannon
Department of Pediatrics
University of California Rm U-432
San Francisco, CA 94143-0724
Tel: (415) 476-7932
Fax: (415) 502-5127

JAPAN

Dr. Haruo Sugano
President Emeritus, Cancer Institute
1-37-1 Kami-Ikebukero, Toshima-ku, Tokyo 170

Dr. Makaoto Goto
Director of Rheumatic Diseases
Tokyo Metropolitan Otsuka Hospital
2-8-1, Minami Otsuka, Toshlma-ku, Tokyo 170

Dr. Akira Horii
Department of Pathology
Tohoku University School of Medicine
2-1 Seiryo-machi, Aoba-ku. Sendai 980-77

Dr. Masao Sasaki
Radiation Biology Center
Kyoto University
Yoshida-konoecho, Sakyo-ku. Kyoto 606-01

Dr. Takehiko Sasazuki
Department of Genetics
Medical Insitute of Bioregulation
Kyushu University
3-1-1 Maidashi, Higashi-ku. Fukuoka 812

Dr. Hiraki Takebe
Department of Radiation Genetics
Faculty of Medicine, Kyoto University
Yoshida-konoecho, Sakyo-ku, Kyoto 606-01

Dr. Kouichi Tatsumi
Division of Biology
National Institute of Radiological Sciences
4-9-1 Anagawa, Inagake-ku, Chiba 263



(2) Workshop on “Lymphoproliferative and Autoimmunal Diseases: Their Reciprocal Relationship.”

A workshop on the question, Lymphoma and Autoimmune Disease: Is There a Reciprocal Relationship, was held at the Hawaii Prince Hotel in Honolulu on February 9-10, 1995. It was organized by Drs. William Blattner (who did not attend) and Robert W Miller, Shunro Sonoda and Haruo Sugano.
Dr. Miller reviewed the several US - Japan workshops previously held on this subject. This workshop will seek to answer two questions:

1. Is there a reciprocal relationship between lymphoma and autoimmune disease in Caucasians as compared with Asians?
   
2. Do differences in the susceptibility to the lymphomagenic vs the neurological effects of HTLV-1 virus serve as a model?

In 1979 Purtilo and Sullivan wrote on the immunological bases for superior survival in females (JAMA 133:1251-3, 1979). They noted that males have higher rates for lymphoma than females do, and that females with systemic lupus erythematosus outnumber males 9:1. Females also have higher rates of other autoimrnune diseases: Graves disease, thyroiditis and Takayasu’s aortitis. They concluded that these and other observations support “the notion that evolutionary selection has equipped females with X-linked immunoregulatory genes for coping with many life-threatening illnesses.” Another observation of interest is that Namba has cited data from 7740 renal transplantations in Japan followed by 5 cases of lymphoma. as compared with 34 among 3823 Caucasians; i.e., 1 in 1500 Japanese as compared with 1 in 100 Caucasians (Kinlen et al, Br Med J 2:1461, 1979 and Japan Society for Transplantation, Ishoku 26:494, 1991). (It is not clear if the immunosuppressive therapy was comparable.)
Stuart C. Finch described in detail the incidence of various forms of leukemia, lymphoma, and multiple myeloma in Caucasians, Japanese, and Japanese-Americans (J-A) in Hawaii and Los Angeles (data for 5 areas in Japan from cancer incidence in five continents). The rates for lymphomas were highest in Japan, lowest in J-A In Los Angeles and intermediary in Honolulu, as if there were an environmental influence. CLL and multiple myeloma had low rates in all three groups.
Autoimmune diseases, more frequent in Japanese than in U.S. whites, are systemic lupus erythematosus, Kawasaki’s disease, Hashimoto’s thyrolditis, Hirata’s disease, autoimmune hepatitis, Takayasu’s aortitis and Behcet’s disease. Less common in Japanese (in order of decreased incidence) are periarteritis nodosa, Crohn’s disease. ankylosing spondylitis, rheumatoid arthritis, diabetes type 1, and ulcerative colitis. Also less frequent are multiple sclerosis, Reiter’s syndrome, and temporal arteritis. About equally frequent in both countries are myasthenia gravis, progressive systemic sclerosis, Sjögren’s disease, and Wegener’s granulomatosis. There is a marked difference in the natural history of myasthenia gravis in U.S. whites vs Chinese, who have few severe cases. In adults with myasthenia gravis 28-38% have thymoma in Japan vs 6-16% in U.S. whites. Some day there may be no autoimmune diseases as they disappear in other rubrics. In multiple sclerosis the HLA haplotypes differ in Japan from those in the U.S.
Shunro Sonoda spoke about ethnic specificity of HTLV-I/II infection and HLA alleles. The African and Asian types differ from one another. Carriers in Japan have similar haplotypes to those in South America and Iran (some of whom have migrated to New York and Europe). Some alleles are pan-ethnic, others are ethnic-specific. Southern Japanese subpopulations are infected with HTLV-I and are at high risk for adult T-cell leukemia/lymphoma (ATL) and HTLV-I-associated myelopathy/tropical spastic paraparesis (HAM/TSP). Their haplotypes are unique and differ from those of the major Japanese population. The unique HLA haplotypes are frequently found among ATL patients and their relatives from southern Japan. The genetic background of Japanese with ATL contrasts with that of patients with HAM/TSP, who are from the major Japanese population.
HLA class I and class II HLA alleles are critically important in distinguishing the genetic make-up of ATL and HAM/TSP. HLA-A26Cw3B61 and DRB1* DQB1* 0901-0303 were typical components of HLA class I and II alleles of ATL patients. The characteristic components of HAM/TSP patients are HLA-A24Cw-B52 and DRB1* DQB1* 0101-0501.
The ethnic specificity of HTLV-I and HTLV-II was studied with regard to the HLA background of native peoples with these infections. HLA-DRB1* DQB1* 0901-0303 and 0403-303 were frequently found in HTLV-I carriers in southern Japan and in natives of the Andes highlands. HLA-DRB1* DQB1* 1402-0301 was specifically found in Orinoco lowland natives in South America and Amerindians in North America. These findings are evidence that susceptibility to HTLV-I and HTLV-II are ethnically determined Environmental factors may still be important.
HTLV-I associated with uveitis, Graves’ disease and other autoimmune disorders was the subject of the presentation of Toshiki Watanabe. Among 166 patients with HTLV-I uveitis 28 had a history of Graves’ disease 2 months to 12 years before. Almost all patients were women. There was no history of Graves’ disease among 222 patients with idiopathic uveitis, who were seronegative for HTLV-I. HTLV-I uveitis has a sudden onset in one or both eyes and a chronic recurrent clinical course. The visual prognosis is good in response to steroid therapy. A main question: is HTLV-1-associated uveitis related to Graves’ disease (both are autoimmune disorders) or to its treatment with methimazole? By in vitro analysis using CAT assay, thyroid hormone was shown to stimulate the promoter activity of HTLV-1 LTR. An 80± base-pair sequence in the U3 region (-141 to -65) is critical for the transactivation. These results suggest that hyperthyroidism can be a precipitating factor for HTLV-I uveitis by increasing the number of infected cells and induce viral gene expression. In addition to HAM/TSP and uveitis there are other inflammatory diseases, apparently of autoimmune etiology. associated with HTLV-I infection: Sjögren’s disease, rheumatoid arthritis and myositis among them.
Norman Talal provided an overview of Sjögren’s syndrome (SS) of dry eyes due to keratoconjunctivitis. dry mouth (salivary gland enlargement) without underlying rheumatic disease, positive for HLA-B8-DR3, and the presence of antinuclear antibodies to Ro/SSA and La/SSB. It is the second most common rheumatic disease. SS may also be secondary to rheumatic arthritis or other connective tissue disease. In most patients significant lymphoproliferation is confirmed to the salivary and lacrimal tissue. In 5-10%, sometimes after 20 years of benign disease, lymphoprollferation extends to the kidneys, lung, lymph nodes, skin, and bone marrow among other sites. SS, an autoimmune exocrinopathy, may develop pseudolymphoma en route to becoming a true monoclonal B-cell lymphoma. Gammaglobulin levels go from hyper to hypo. The pathogenesis involves a genetic disorder influenced possibly by sex hormones, stress and a virus. Interest centers on CD5+ B cells, which circulate in chronic lymphocytic leukemia (CLL). They have been associated both with autoimmunity and with CLL in mice and humans. Five percent of SS patients develop MALT lymphomas which may be low grade or rapidly progressive. Lymphocytes infiltrating the salivary glands fail to undergo apoptosis, perhaps because bcl-2 is also expressed in these cells. This inhibition of apoptosis may be a common theme in autoimmunity and lymphoma.. Finally, salivary gland lymphoid infiltrates and parbtid swelling can occasionally be part of the spectrum of clinical and laboratory features associated with HIV infection. Because the clinical findings may resemble those of SS, HIV must now be considered in its differential diagnosis.
Steven Jacobson spoke on cytotoxic T cell responses in patients with HTLV-I associated neurological disease. The spinal fluid and peripheral blood of these patients have high levels of circulating HTLV-I specific CD8+, HLA class 1 restricted cytotoxic T cells (CTL) that are known to be restricted to immunodominant regions of the HTLV-I tax protein. There are low levels or none in asymptomatic individuals who are seropositive for HTLV-I. Specific peptides of the HTLV-I tax protein can be recognized in the context of a particular class of HLA class I alleles. An estimate of the frequency of a precursor of these HTLV-I tax-specific cytotoxic T cells indicated a very high frequency in both peripheral blood lymphocytes and cerebrospinal fluid: 1 in 75 to 1 in 280. These cytotoxic cells, it was thought, may contribute to the development of neurological disease associated with HTLV-I, and may be a target for attempts to eliminate them through immunotherapy. The usage of T cell receptors by these HTLV-I specific cytotoxic T cells has been characterized in a patient with early HAM/TSP who exhibited preferential use of V!!!and V!!!T cell receptor chains. Moreover, the CDR3 region of the dominant T cell receptor V!!!chain was identical in three patients of the same HLA class 1 haplotype. It is hoped that better understanding of the mechanisms of HTLV-I-associated neuropathology can be gained by analyzing such antigen-specific functional cellular host responses to HTLV-I. Based on these models, it may be possible to devise immunotherapeutic strategies for disease intervention.
Angela Manns said that HTLV-I and its related diseases, adult T-cell leukemia/lymphoma (ATL) and HTLV-I associated myelopathy/tropical spastic paraparesis (HAM/TSP) exemplify the reciprocal relationship between a lymphoproliferative and an immune disorder. This relationship demonstrates that a common etiologic agent can cause different disease manifestations in disparate geographic locales. In the Caribbean. particulariy Jamaica and Trinidad, the NCI Viral Epidemiology Branch has collaborated with the University of the West Indies in studies that are elucidating common and distinct features of HTLV-I infection, ATL and HAM/TSP as compared with HTLV-I endemic areas of Japan. The Caribbean, like Japan, has low rates of certain lymphoproliferative disorders with high rates of various autoimmune disorders. ATL has a higher incidence, 4 per 100,000 as compared with 3 per 100,000 in the Caribbean. At diagnosis ATL patients are ten years older in Japan. which suggests that environmental or other co-factors differ in the two locales. However, the lifetime risks for ATL of 1-5% among HTLV-I carriers are similar. Her group’s epidemiologic approach has been to characterize virus exposure, immunology and the genetics of ATL and HAM/TSP patients. Through descriptive and analytical studies, differences have been demonstrated in acquisition of the virus through various routes of transmission. ATL is linked to early life exposures from maternal to infant transmission via breast-feeding, whereas HAM/TSP is linked either to parenteral or sexually acquired disease. Contrasts have been identified in laboratory parameters between the two diseases which support the existence of distinct immunogenetic pathways. For example, cytokine profiles representing cell-mediated immune response for HAM/TSP reveal elevated interferon gamma, tumor necrosis factor-alpha, interleukin-1-beta consistent with an inflammatory response whereas ATL has a response consistent with immune deficiency revealing elevated levels of transforming growth factor-beta Differences in the immunogenetic background of ATL and HAM/TSP patients have also been identified, and may in part account for the differential development among HTLV-I carriers of either ATL or HAM/TSP. The application of epidemiologic techniques to the evaluation of the reciprocal relationship between lymphoproliferative and autoimmune disease can provide insights into their common and diverse pathogenic mechanisms.
Katsuyuki Aozasa and his associates reported that 3 of 134 patients with chronic pyothorax developed pyothorax-associated lymphoma (PAL) in one hospital for chest diseases in Osaka. No cases of PAL, however, could be found in over 1600 cases of malignant lymphoma in general hospitals in the same district, which may indicate that pyothorax played a role through chronic stimulation by an autoimmune reaction. In comparing 42 patients with PAL and 70 with pyothorax but no lymphoma. the number of PAL who had had artificial pneumothorax was 5 times greater than in the controls (p<0.05). A nationwide survey was made and 37 cases of PAL were found, whose chronic pyothorax ranged from 22-55 years. 81% had tuberculosis, 16% with pleurisy. All tumors were lymphocytic lymphoma, with diffuse large cell type being the most common. PCR and In situ hybridization together with LMP-1 immunohistochemistry revealed 85% of PAL were associated with EBV. Cases have been reported in Asia. but rarely in Western countries.
Masako Tanimura spoke on lymphoma/1eukemia in children with severe immunodeficiency, almost all of it inborn. Among 29,804 cases in the Japan Children’s Cancer Registry, started in 1969, 12 were known to be immunodeficient, most detected from a systematic search begun in 1984. Five had ataxia-telangiectasia (AT): 3 lymphatic leukemia, 1 Burkitt’s lymphoma, and l non-Hodgkin’s lymphoma. Two had IgA nephropathy: 1 medulloblastoma and 1 papillary thyroid cancer. Chediak-Higashi syndrome and Bloom’s syndrome each were associated with malignant lymphoma. There was 1 each of IgA deficiency and Letterer-Siwe disease, hyper IgE with retino-blastoma, and hypogammaglobulinemia with ALL.
The other source was an Immunodeficiency Registry started in 1974. Of 642 patients under age 15 with primary immunodeficiency, 19 had malignant tumors. Only 2 cases were found in both registries. Among 56 children with AT, 7 had cancer: 3 malignant lymphomas, 2 ALL, 1 gastric cancer and 1 hepatoma. The other diagnoses were SCID (1 Letterer-Siwe disease, 1 lymphoma), and IgA deficiency (1 Hodgkin’s disease, 1 glioma). Six others had lymphoma: 2 Chediak-Higashi syndrome, 2 Wiskott-Aldrich syndrome, 1 CVID, and 1 hyperIgE syndrome. Also l child with hyperIgM had a brain tumor, and another with transient hypogammaglobulinemia had ALL.
Carlo M. Croce reviewed the genetic mechanisms of lymphomagenesis and leukemogenesis as a prelude to describing self-fusion of the ALL1 gene, a new genetic mechanism for acute leukemia. One mechanism leads to activation of oncogenes by reciprocal chromosomal translocation. Burkitt’s lymphoma develops when the c-myc gene on chromosome 8 is activated by a translocation to chromosome 14, 2 or 22. A second mechanism is gene fusion, in which a chromosomal rearrangement brings two different genes into juxtaposition to create a composite of the head of one gene and the tail of another. The result is the encoding of a chimeric protein in which both genes contribute to the oncogenic properties of the protein. Ninety percent of persons with chronic myelogenous leukemia (CML) have a fusion of the bcr gene, normally at chromosome 22q11 and the c-abl gene on 9q34. The resulting bcr/abl gene encodes a protein whose high tyrosine kinase activity contributes to expansion of the neoplastic myeloid clone. A newly delineated third genetic mechanism is self-fusion in which there is no microscopic evidence of a translocation in certain leukemias. The ALL1 gene, which can arise by cell fusion as just described, can also arise by self-fusion, in which the gene is rearranged by duplication of an internal portion of the gene. This partial duplication fuses the ALL1 gene to a portion of itself. Southern blot analysis has detected self-fusion (within chromosome 11) in 2 of 19 adults with acute myelogenous leukemia (AML) without translocation and in 3 of 4 AML patients with trisomy 11. Their AML was of the M1 or M2 type. In 2 patients with trisomy 11, exon 6 was fused with exon 2. Trisomies of other chromosomes in leukemic patients should be studied for this new mechanism. Molecular probes are being developed to screen for ALL1 in leukemia under 1 year of age (in whom 95% have a transiocation of chromosome 11), therapy-related AML, myelodysplastic syndrome, monocytic morphology or bilineage phenotype. A report of these findings was published the following week (Schichman SA et al: JAMA 273;571-576, 1995),
Nanao Kamada opened his presentation by responding to the questions put to each participant before the workshop. Is there a reciprocal relationship between the frequency of lymphoma and autoimmune diseases? He did not think so because a relatively small series of lymphoma after immunosuppression for organ-transplantation did not seem much different in Japan (Yokota 1990) than in Australia (Sheil, 1979) (see Miller above for dissimilar results in a large series) Also the occurrence of 8 lymphoma in 23 Japanese patients with AIDS and cancer were extranodal with EBV genome did not suggest decreased susceptibility to lymphoma. In the U.S. 60% are extranodal. The CD4/CD8 ratio in Japan is not different from that found elsewhere (Fujimoto 1985, Kobayashi 1986). He answered yes to the question, is HTLV-I a good model to study the relationship between lymphoma and autoimmune disease because of the severe immunological disturbances induced by the infection. He then went on to discuss molecular-cytogenetic findings in l) prelymphoma and 2) HTLV-I related lymphoma. The study concerned 11 cases of one form of prelymphoma, angioimmunoblastic lymphadenopathy with dysproteinemia (AILD) and 107 shouldering adult T cell leukemia/lymphoma (smATLL). The focus was on the clonality of abnormal cells in AILD and SmATLL, using cytogenetic, molecular genetic and FISH data. A schematic diagram showed a clinical comparison of the two disorders, and the immunological changes as the diseases progress. HTLV-I infection progresses through and intermediate stage (polyclonal Integration) which may lead to B-cell lymphoma, but generally leads to smoldering ATL (skin eruption, increase of CD4/CD8 and sometimes with hypergarnmaglobulinemia) and then on to overt ATL. AILD (no HTLV-I infection) may also lead to B-cell lymphoma, but typically ends with T-cell lymphoma, after passing from hyper-gammaglobulinemia to hypogammaglobulinemia with intermediate skin eruption sometimes, positive Coombs test, allergy to drugs, increase in the CD4/CD8 ratio, and the production of autoantibodies.
Susumu Ikehara reported on his group’s many studies of autoimmune diseases as stem-cell disorders in mice. First they showed that transplantation of the thymus from a mouse with autoimmune disease to a normal mouse did not cause disease autoimmune disease, but bone-marrow transplantation did. The same effect was produced by autoimmune-disease stem-cell concentrates. Among murine diseases that have been induced, prevented or treated by stem cells are diabetes mellitus (insulin-dependent or not), immune thrombocytopenic purpura, segmental glomerulosclerosis, and systemic lupus erythematosus. When bone marrow donors harbor one of these diseases, the recipients has developed it.
These observations provide direct evidence that the origin of autoimmune diseases, either organ-specific or systemic, is attributable to defects in the hematopoietic stem cells, which produce polyclonal abnormal stem cell proliferations. This knowledge is leading to effective treatment of these diseases. Ikehara’s group has shown that three types of cells are essential to successful marrow transplantation: pluripotent hematopoietic stem cells, natural suppressor cells and stromal cells present in bone marrow. He has noted that about half of the 86 intractable diseases designated by the Japanese Ministry of Health and Welfare, are curable by bone marrow transplantation. Overviews have been published recently by Ikehara (Pathol Internatl 44:817-26, 1994), and Ikehara et al (Tohoku J Exp Med 173:141-55, 1994)
Haruo Sugano reviewed the role of Epstein-Barr virus (EBV) in gastric medullary carcinoma with marked lymphoid infiltration. By the PCR method, the presence of EBV-DNA was detected in 28 of 30 cases, and in 27 (90%) EBV infection limited to cancer cells was identified by in situ hybridization with the EBER probe. Monoclonal EBV infection was found in 70%. Only one case, however, was positive for anti-LMP-1 antibody, and none bound anti-EBNA-2 antibodies. The role of EBV in this particular form of gastric cancer is obscure. It Is histologically similar to nasopharyngeal carcinoma, well known to be associated with EBV. In addition, the detection of the EBV genome has also been reported in similar histologies in the lung, thymus, and salivary gland. Recently the etiologic role of EBV has been of interest in Hodgkin’s disease, lymphoma of the stomach, leiomyosarcoma in children with AIDS, and smooth-muscle tumors occurring after organ-transplantation. By contrast, EBV genome was detected in about 10% of usual gastric cancer and 21% of breast cancer. In summary, the role of EBV in lymphomagenesis as well as in tumorigenesis should be reconsidered in relation to immune surveillance and immunodeficiency.
At the workshop we did not take special note that Dr. Finch showed that in Japan the rates of one-third of autoimmune diseases were high, one-third are low and one-third are the same as compared with the U.S. Also, the rates of various autoimmune diseases were unequally high or low. Thus the relationship is not reciprocal. Dr Sonoda showed that HTLV-I and II diseases vary in susceptibility among national and international subgroups in relation to HLA haplotypes, and lymphoma occurs in some, whereas HAM/TSP (autoimmune) occur in others. Dr. Watanabe noted that HTLV-I is associated with autoimmune diseases (uveitis, Graves’ disease and others). Dr. Talal said that Sjögren’s is an autoimmune disease followed in some cases by lymphoma; i.e., both diseases in the same patient, as gammablobulin levels go from too high to too low. He stated that blocking apoptosis in T lymphocytes leads to Sjögren’s disease, rheumatoid arthritis or SLE. Dr. Jacobson stated that cytotoxic T cells may contribute to the development of HAM/TSP, and immunotherapy may hold promise. Dr. Aozasa’s studies of pyothorax-associated lymphoma, 85% of which had EBV infection, may be due to chronic stimulation that leads to an autoimmune reaction. (Note that in several instances lymphoma is ascribed to autoimmune disease.) Dr. Tanimura presented data on the high frequency of lymphoma in immunodeficiency disorders [which are late occurrences in AILD (Dr. Kamada said) and Sjögren’s disease]. Dr. Croce referred to the chromosomal-immunologic abnormality in Burkitt’s and the role of gene fusion and self-fusion in the pathogenesis of ALL1-associated leukemias.

PARTICIPANTS

UNITED STATES

Dr. Carlo M. Croce
Thomas Jefferson University
233 South 10th Street, Room 1050
Philadelphia, PA 19107

Dr. Stuart C. Finch
Vice-President for Research and Development
Cooper Hospital University Medical Center
Sarah Cooper Building, Room 118
1 Cooper Plaza
Camden, New Jersey 08103-1489

Dr. Steven Jacobson
National Institute of Neurological Disorders and Stroke
Building 10, Room 5B10
Bethesda, Maryland 20892

Dr. Angela Mauns
National Cancer Institute
Viral Epidemiology Branch
EPN 434
Bethesda, Maryland 20892

Dr. Robert W Miller
National Cancer Institute
Scientist Emeritus
Clinical Epidemiology Branch
EPN 400
Bethesda, Maryland 20892

Dr. Norman Talal
University of Texas
Health Science Center
7703 Floyd Curl Drive
San Antonio TX 78284

JAPAN

Dr. Haruo Sugano
Director Emeritus
Cancer Institute
1-37-1 Kami-Ikebukuro Toshima-ku
Tokyo 170

Dr. Katsuyuki Aozasa
Department of Pathology
Osaka University School of Medicine
2-2 Yamadaoka Suita Osaka 565

Dr. Susumu Ikehara
Department of Pathology
Kansai Medical University
10-16 Fumizono-cho, Koriguchi, Osaka 570

Dr. Nanao Kamada
Department of Cancer Cytogenetics
Research Institute for Radiation, Biology and Medicine
Hiroshima University
1-2-3 Kasumi Minami-ku Hiroshima 734

Dr. Shunro Sonoda
Department of Virology
Kagoshima University, Faculty of Medicine
8-35-1 Sakuragaoka, Kagoshima 890

Dr. Masako Tanimura
Head of Developmental Psychology
Research Laboratory
National Children’s Medical Research Center
3-35-31 Taishido, Setagaya-ku
Tokyo 154

Dr. Toshiki Watanabe
Department of Pathology
Institute of Medical Science
University of Tokyo
4-6-1 Shirokanedai, Minato-ku, Tokyo 108