Familial Aggregation of Cancer from Proband Cases with Childhood Adrenal Cortical Carcinoma1
Yukiko Tsunematsu,2 Shaw Watanabe,3 Toshiaki Oka,4 Taiji Tsukamoto,5 Keisei Kawa-Ha,6 Yoshiaki Hirata,7 Hiromichi Yamanaka,8 Mutsuro Ohira3 and Michiko Ono9
2 National Children’s Hospital, 3-35-31 Taishido, Setagaya-ku, Tokyo 154, 3National Cancer Center Research Institute, 5-1-1 Tsukiji, Chuo-ku, Tokyo 104, 4Asahikawa Medical College, Nishikagura, Asahikawa, Hokkaido 078, ‘Sapporo Medical College, Minami 1-jo, Nishi 17-chome, Chuo-ku, Sapporo, Hokkaido 060, 6Osaka University Hospital, 1-1-50 Fukushima, Fukushima-ku, Osaka 553, ‘Hamamatsu Medical Center, Hamamatsu, Shizuoka 431, 8Aichi Medical College, Nagakute, Aichi- gun, Aichi 480-11 and 9Tokyo NTT Hospital, Fujimi, Chiyoda-ku, Tokyo 102
Family pedigree of Li-Fraumeni syndrome was investigated from probands with childhood adrenocor- tical carcinoma in Japan. From 47 probands, 7 families had 3 or more cancer cases at ages less than 45 years within the first generation; one satisfied Li’s original criteria, two were acceptable because of multiple primary cancer in the probands, and others showed an aggregation of cancers with onsets at early ages, though no sarcoma of mesenchymal origin was found. A significantly higher occurrence of cancer in the mothers of the probands, especially of the breast, was consistent with reports from the USA, and liver cancer, osteosarcoma and lung cancer among family members under the age of 45 also showed a higher frequency than in the general population. Similarities and differences between Japanese and Caucasian cases are discussed.
Key words: Li-Fraumeni syndrome - Adrenocortical cancer - Breast cancer - Epidemiology
Studies of familial aggregation of cancer have con- tributed to clarification of the genetic influence upon cancer occurrence.1,2) Childhood cancer is a good target for genetic study, because the environmental exposure which often hides the effect of genetic or individual susceptibility is considered to be less than that in adults. Hereditary retinoblastoma is a good example of how dominant inheritance is caused by the deletion of the RB gene, which could be related directly to carcinogene- sis.1,3) Loss of heterozygosity of the RB gene has been found to be related to many other cancers, and now RB is recognized as one of the important suppressor genes.4)
In Li-Fraumeni syndrome and so-called Lynch’s sar- coma, breast cancer, lung cancer and adrenocortical cancer (SBLA) syndrome, a complicated accumulation of childhood and adult cancers in one family is usually found, and it is of interest to establish whether or not these various cancers may be caused by a single gene or multiple gene events.5-8)
The criteria for Li-Fraumeni syndrome and SBLA syndrome have not yet been precisely defined, because of the rather broadly defined concept including various tumors. An alternative name of sarcoma-breast cancer syndrome has been proposed, but it has not yet been widely accepted.2) Frequency and incidence of this syn-
drome in different ethnic groups or different countries have not yet been clarified. We tried to confirm whether or not families with Li-Fraumeni syndrome are present in Japan. As the proband, we chose childhood adrenal cortical cancer, because this carcinoma is involved in both Li-Fraumeni and SBLA syndrome and its rarity could exclude cancer aggregation by chance.9, 10)
MATERIALS AND METHODS
Childhood adrenocortical cancer cases were selected from the Children’s Cancer Registry of Japan (Director: Dr. Noboru Kobayashi), and were found in the literature by a survey of the Japan Information Center of Science and Technology (JICST) data base. The Children’s Cancer Registry of Japan covers about half of the inci- dent cases of childhood cancers in Japan. Registered numbers of childhood cancer were 5,666 during 1969- 1973, 5,929 during 1974-1978, and 6,785 during 1979- 1983.11) The JICST data base is maintained by the Agency of Science and Technology and contains more than 1.5 million references in medicine.
Thirty cases were found in the registry between 1969 and 1984, and 24 cases by JICST between 1965-1988. Seven cases were duplications. All cases in the literature were reported by clinicians, and two out of 24 were cases with multiple primary cancer; others were solitary cases reported mainly because of their rarity. The number of probands reported to the registry was 4 during 1969-
’ A part of this work was presented at the 81st Annual Congress of the American Association of Cancer Research in 1990.
2 To whom correspondence should be addressed.
1973, 6 during 1974-1978, 15 during 1979-1983, and 5 in 1984. The total number of probands was 47 after exclu- sion of duplicated cases. The numbers of cases in the literature were 3, 6, and 5 respectively, corresponding to the above three periods, and 3 during 1984-1988.
Diagnosis of adrenocortical carcinoma of probands had been confirmed by surgical pathology and/or autopsy. Pathology reports were reviewed, if possible, but a review of glass slides was not done.
Questionnaires were sent to the attending physicians in June 1989 to confirm the family pedigrees by collecting new incident cases of cancer since the first registry or report. Thirty-five out of 47 physicians (74.5%) responded, and up-to-date follow-up information was obtained from 25 of them. Telephone interviews were conducted for 22 pedigrees and present status was con- firmed in 15 families. Family pedigrees of 12 non- respondants were used from the initial registry or report. The number of families providing follow-up information according to source is shown in Table I.
The proportions of histological types of cancers appearing under the age of 45 among family pedigrees were compared with those among the Japanese popula- tion and the US white population. 12, 13) Proportional inci- dence ratios for major cancers were calculated using the sex- and age- (0-44 years old) specific proportions of each cancer site. Confidence intervals (95%) for proportional incidence ratios were calculated according to the method of Breslow and Day.14)
Frequency of cancer occurrence in the mother was compared separately with expected cancer occurrence calculated from the Japanese cancer incidence by site and sex.12) The mothers’ ages had not been filled out in 5 registry cases and 7 reported cases, so 35 mothers’ ages were available for calculation. Observed/expected ratio was tested by Breslow’s method.14) Cumulative occur- rence of cancers in mothers of 35 probands was calcu- lated and tested by the Kaplan-Meier method.
RESULTS
The characteristics of the probands were as follows: male-to-female ratio of probands was 1:1.5, and average age at onset was 4.4±3.2 in boys and 4.5±3.0 in girls. Age distribution is shown in Fig. 1. Eleven proband cases were alive, 26 dead and 10 were lost to follow-up. Multi- ple cancer in the probands was recognized in 2 cases. One had triple primary cancer of synchronous adrenocortical carcinoma and neuroblastoma at age 7 months and rhabdomyosarcoma at age 4 years,15) and another had adrenocortical carcinoma at age 11 months and osteo- sarcoma at age 15 years. Two cases showed hemihyper- trophy and one showed overgrowth (+3.5 SD in height and bone age being 10 years at 5 years of age).
The numbers of responses to questionnaires and tele- phone interviews by the source of probands are shown in Table I. Follow-up information was obtained from 25 out of 30 probands in the registry series, and 10 out of 17 in the literature series. Information about cancer incidence was obtained for at least two generations.
No. of cases
5
Males
0
5
10
15
Age
5
Females
0
5
10
15
Age
| No. of proband | Response + | Follow-up information | Additional cancer <45 yrª) | Further additional cancerb) | |
|---|---|---|---|---|---|
| Registry | 30 | 25 | 17 | 6 | 3 |
| Literature | 17 | 10 | 9 | 3 | 1 |
| Total | 47 | 35 | 26 | 9 | 4 |
a) <45 yr, younger than 45 years of age.
b) Data obtained by telephone interview.
| Sarcoma | Brain | Lymphoma | Breast | Lung | Stomach | Liver | Uterus | Ovary | Others | Total | |
|---|---|---|---|---|---|---|---|---|---|---|---|
| Proband | 2* | 1 * | 3 | ||||||||
| Sibling | 1 | 1 | 2 | ||||||||
| Half sib | 1 | 1 | 2 | ||||||||
| Father | 1 | 2 | 1 * | 4 | |||||||
| Mother | 3 | 1 | 1 | 1 | 6 | ||||||
| Grandfather | 4 | 1 | 5 | ||||||||
| Grandmother | 2 | 1 | 1 | 3 | 7 | ||||||
| Uncle | 1 | 1 | 1 | 3 | 6 | ||||||
| Aunt | 2 | 2 | 1 | 5 | |||||||
| Cousin | 1 | 2 | 3 | ||||||||
| Total | 4 | 2 | 1 | 7 | 4 | 7 | 4 | 5 | 2 | 7 | 43 |
* Multiple primary cancer cases.
| < 15 yr | 15-44 yr | >45 yr | |
|---|---|---|---|
| Sarcoma | 4 | - | |
| Brain tumor | 1 | 1 | |
| Lymphoma | 1 | ||
| Breast | 6 | 1 | |
| Lung | 2 | 2 | |
| Stomach | 5 | 2 | |
| Liver | 3 | 1 | |
| Uterus | 2 | 3 | |
| Ovary | 2 | ||
| Others | 1 | 5 | 1 |
| Total | 10 | 24 | 9 |
The number of cancers among relatives had been 18 at the time of initial diagnosis of proband cases, but 27 additional incident cases were found by this survey. Thirty-four out of 43 (79%) occurred at the age of 45 or under. Frequently occurring cancers were of the breast, stomach, uterus and liver (Table II). Other cases in- cluded 1 colon cancer in a sibling, multiple cancer of the submandibular gland and lung in the father, rectum cancer in the grandfather and one each of cancer of the pancreas, bladder and nasopharynx in uncles.
The ages at diagnosis for these cancers in relatives were younger even for the more common cancers, such as of the stomach and liver (Table III). Two of the four liver cancers were hepatoblastoma. The proportions of these cancers were compared to those among the general populations in Japan and the USA (Table IV). A significantly higher proportional incidence of cancer was recognized in osteosarcoma and liver, and a higher fre- quency was suggested in the breast, brain and lung.
Breast cancer among mothers was further analyzed. Six mothers developed cancer (3 breast, 1 stomach, 1 lung and 1 ovary). Observed/expected ratio of mothers’ cancer (all sites) was 15.4 (99% confidence interval (CI); 3.9-40.3), and O/E ratio of mothers’ breast cancer was 50.3 (99% CI; 5.1-184.0). Cumulative probability of mothers’ cancer occurrence in comparison with the gen- eral population in Japan is shown in Fig. 2.
Analysis of each family pedigree revealed that 7 families had 3 or more cancers within the first generation (Fig. 3). Only one family satisfied the original criteria of Li, in which 3 were childhood cancer cases including lymphoma and father’s lung cancer at a young age of onset (Fig. 3a). The father of the proband case had married twice, and 3 of 4 children from two mothers had childhood cancers. Three of the father’s 6 siblings had died of cancer, one of osteosarcoma. If the multiple primary cancers in the proband were counted as two cancers, two other families would be classified as having Li-Fraumeni syndrome (Fig. 3b, c). Four other families had 3 or more cancers within the first generation, but these families were characterized by a lack of sarcoma of mesenchymal origin (Fig. 3d-e).
DISCUSSION
Li-Fraumani syndrome resembles Lynch’s SBLA syn- drome in many respects, but the distribution of cancer is somewhat different (Table V). Li6) set the criteria that 1) the proband may have any sarcoma which occurs at age less than 45, 2) among first-degree relatives, at least two of any cancers occur at less than 45 years of age, or 3) among first- or second-degree relatives, any cancer occurs at less than 45 years of age or sarcoma at any age.6) He later revised the criteria to be that sarcoma at less than 45 years of age in the proband and cancers in
| This study | Japan | PIR (CI) | USA | ||
|---|---|---|---|---|---|
| No. | % | 1985 | 1980-1985 | ||
| Soft tissue | 1 | 2.9 | 1.2 | 2.41 (0.06-13.65) | 2.0 |
| Bone | 3 | 8.8 | 0.9 | 9.77* (2.02-28.62) | 1.2 |
| Leukemia | 0 | 0 | 6.7 | 5.5 | |
| Lymphoma | 1 | 2.9 | 3.3 | 0.88 (0.02-4.96) | 9.4 |
| Brain tumor | 2 | 5.9 | 4.2 | 1.40 (0.17-5.05) | 5.0 |
| Stomach | 5 | 14.7 | 23.1 | 0.64 (0.21-1.49) | 1.0 |
| Colon | 1 | 2.9 | 4.1 | 0.71 (0.02-4.01) | 3.2 |
| Liver | 4 | 11.8 | 2.9 | 4.07* (1.10-10.39) | 0.6 |
| Pancreas | 1 | 2.9 | 1.1 | 2.64 (0.07-14.87) | 0.8 |
| Lung | 2 | 5.9 | 3.1 | 1.90 (0.23-6.86) | 5.2 |
| Breast | 6 | 17.6 | 13.8 | 1.28 (0.47-2.79) | 20.8 |
| Others | 8 | 23.5 | 35.5 | 0.66 (0.28-1.30) | 45.0 |
| Total | 34 | 100.0 | 100.0 | 100.0 | |
Proportion of cancer was calculated from Japanese incidences of cancer12) and that of the SEER program13) in 1985. PIR, proportional incidence ratio; CI, confidence interval; * , P<0.05.
%
60
All Sites
40
Probability
20
Breast
All Sites
Breast
0
25
30
35
40
45
50
55
60
65
Mother’s Age at Last Follow-up
at least two young relatives should be present.16) Breast cancer frequently occurring at a young age of onset and bilaterally, and osteosarcoma, brain tumor, leukemia, adrenocortical cancer recognized in one family would satisfy the revised criteria. He also accepted that multiple primary cancers may be equivalent to cancers in two members of a family. SBLA syndrome includes sarcoma, breast cancer, brain tumor, lung and laryngeal cancer, leukemia and adrenocortical carcinoma in family pedi- gree. Multiple primary cancer is also frequently found.8)
We started with adrenocortical carcinoma as the pro- band, because it is included in both syndromes, and leukemia and sarcoma occur too commonly as tumors in childhood, which would dilute the genetically related accumulation of cancer. It is difficult to distinguish genetically determined cancer from common cancer if the proportion of inherited cancer is small. The frequency of adrenocortical carcinoma among all cancers at less than 45 years of age from the Annual of the Pathological Autopsy Cases in Japan (1975-1988) was 74 out of
a
b
I
I
.
4
1
4
70
65
70
63
d57
St
B57
II
#
·
II
I
II
Lu35
+ Submand, Gl, Ca.35
39
40
37
38
54
B
₫45 Lu45
d5
39
45
Os 5
III
13
II
e
₫4
17
16
AC 7m+ Neur. 7m
HI
1
5
Rhab. 4
AC 5
Os
d 10
₫5
11
Lymphoma
Hepatoblastoma
C
I
I
4
d 65
76
48
II
¢
8
51
48
42
57
54
51
d43
842
III
4
d
16
14
22
d 19
1
1
I
AC Os
| Im.
15
d 33
B 33
e
1
I
4
UL
II
77
32
65
62
60
142
Po42 B48 Ov24
d48
d25
·
13
III
I
36
V34 d14 2
28
36
d18
32
30
29
60
57
53
48
46.
42
Hap42
Lu44
51
49
47
45
Ur 34
83
35
Ov18
42
IV
A
19
A65
3
M
C
2
BT9 (medullob.)
AC 2
26
AC 5
1
f
g
I
I
4
I
17
d41
77
St St 30 33
S141
II
10
II
19
69
52
039
d37
d38
45
45
d32
Px39 S137 835
B32
d36 ₫20 43 39
37
Pa BI BT
36 20 4!
(Astrocytomo)
IK
U
2
-
III
1
7
d20
F08
26 23 12
17 15 13
d 5
15
8
C20
AC8
Hepotoblastoma 2
AC 4
| Li-Fraumeni syndrome | Lynch's SBLA syndrome | |
|---|---|---|
| Proband | Any sarcoma <45 yr | |
| Criteria | Cancers in at least | |
| 2 young relatives | ||
| Frequently occurring tumors | Breast, often at a young | Sarcoma |
| age and bilateral | Breast | |
| Osteosarcoma | Brain | |
| Brain tumor | Lung | |
| Leukemia | Larynx | |
| Adrenocortical cancer | Leukemia | |
| Adrenocortical cancer | ||
| Multiple primary cancer | Multiple primary cancer |
37,749 (0.20%), and there were only 17 cases under the age 15 years.17)
The number of pedigrees which had more than two cancers in the first generation in addition to the proband case, occurring at less than 45 years of age, was 7 out of 47 (14.9%). Although 18 cases of cancer in the family pedigree were found among 47 families (38.3%) at the time of initial diagnosis of the proband case, twenty- seven new cancer cases (77.1% of total cancers among pedigrees) were found in the follow-up, demonstrating the importance of this kind of study.18)
Penetration or expression rate of cancer gene(s) in Li-Fraumeni syndrome has not yet been clarified. Li’s 24 families had been collected from cases in the Cancer Family Registry of the Epidemiology Branches at NCI, so the selection bias yielded a seemingly high penetration rate.6) In this series, it seemed to be lower than that reported in the USA. Only one family is acceptable to be classified as Li-Fraumeni syndrome by the original criteria, and two others may be acceptable if multiple primary cancer of the proband may be counted as two. In these two family pedigrees, however, cancer was only present in the mother and child or father and child. Four other family pedigrees included more than 3 younger onset cancers within one generation, but a lack of sar- coma was characteristic.
Breast cancer was the cancer most frequently asso- ciated with childhood sarcoma.19-23) In our series, how- ever, 3 mothers out of 47 (6%) developed breast cancer. Four others were breast cancer in the grandmother or aunts. Although the incidence of breast cancer in Japan was about one-fourth of that in the USA in this age category,24) the proportional cumulative incidence by age 45 was 1.6 times higher than that of the Japanese general population.
The frequency of adrenocortical carcinoma in Li- Fraumeni’s series was about 10%. If this frequency could
be applied to data from Japan, the number of families with genes related to this syndrome may be 10 times greater. On the contrary, family pedigrees starting with probands with sarcoma or leukemia could not be detected in 4 of these families. Four families showed a combination of rather common epithelial tumors, such as stomach, pancreas and colon cancer, at a young age of onset, in addition to brain and/or germ cell tumors. The validity of including germ cell tumors in Li-Fraumeni syndrome was proposed by Hartley et al.25) Four of these families gave us the impression that they might belong to the intermediate type of Lynch’s cancer family syndrome type I and type II. The alternative name of sarcoma- breast cancer syndrome, instead of Li-Fraumeni syn- drome or SBLA (Lynch’s type II) syndrome, seems to be somewhat premature, because of the presence of families such as those above and the low frequency of breast cancer in countries like Japan.
The mechanism of multiple occurrence of different cancers according to dominant heredity is not known. At present there is no consistent genotypic or phenotypic marker to identify individuals who carry this putative cancer-predisposing gene. Little et al.26) examined radia- tion resistance of 6 cell strains from one family represent- ing affected, nonaffected, and at-risk individuals and two strains from affected individuals from two other families, and found no difference in cytotoxic effects or production of chromosome aberrations by radiation.
Yano et al.27) reported a loss of heterozygosity in hu- man adrenocortical carcinoma at 17p (6/6), 11p (4/6), 13q (3/6), and one HRAS1 mutation (1/9). As benign tumors and hyperplasia did not show such chromosomal deletions (0/8), accumulated changes at tumor suppres- sor loci in particular organs could lead to the different histological types. Inactivation of several tumor suppres- sor genes and activation of oncogenes is considered to be essential for the development of malignancy.28) Malkin et
al.29) recently reported that point mutation is present in the p53 tumor suppressor gene in lymphocytes and/or fibroblasts of the patients of Li-Fraumeni syndrome. Two of our cases also had p53 gene alteration (Yokota et al., in preparation). Further linkage studies on families with DNA changes would be desirable from a molecular epidemiological viewpoint.
ACKNOWLEDGMENTS
A part of this research was supported by Grants-in-Aid for Cancer Research and for the Comprehensive 10-Year Strategy for Cancer Control from the Ministry of Health and Welfare. We thank Drs. F. Li, L. Strong and N. Kobayashi for their support and constructive suggestions during this work.
(Received February 22, 1991/Accepted May 15, 1991)
REFERENCES
1) Knudson, A. G., Jr. Retinoblastoma: a prototypic hered- itary neoplasm. Semin. Oncol., 5, 57-60 (1976).
2) Miller, R. W., Li, F. P. and Sugano, H. The family syndrome of sarcoma, breast cancer, and other neoplasms: meeting report. Jpn. J. Cancer Res., 79, 1155-1158 (1988).
3) Cavenee, W. K., Dryja, T. P., Phillips, R. A., Benedict, W. F., Godbout, R., Gallie, B. L., Murphree, A. L., Strong, L. C. and White, R. L. Expression of recessive alleles by chromosomal mechanisms in retinoblastoma. Nature, 305, 779-784 (1983).
4) Lee, E. Y., To, H., Shew, J-Y., Bookstein, R., Scully, P. and Lee, W. H. Inactivation of the retinoblastoma suscep- tibility gene in human breast cancers. Science, 241, 218- 221 (1988).
5) Li, F. P. and Fraumeni, J. F., Jr. Soft-tissue sarcomas, breast cancer, and other neoplasms. A familial syndrome? Ann. Intern. Med., 71, 747-752 (1969).
6) Li, F. P., Fraumeni, J. F., Jr., Mulvihill, J. J., Blattner, W. A., Dreyfus, M. G., Tucker, M. A. and Miller, R. W. A cancer family syndrome in twenty-four kindreds. Cancer Res., 48, 5358-5362 (1988).
7) Birch, J. M., Hartley, A. L., Blair, V., Kolsey, A. M., Harris, M., Teare, M. D. and Jones, P. H. M. Identifica- tion of factors associated with high breast cancer risk in the mothers of children with soft tissue sarcoma. J. Clin. Oncol., 8, 583-590 (1990).
8) Lynch, H. T., Katz, D. A., Bogard, P. J. and Lynch, J. F. The sarcoma, breast cancer, lung cancer, and adrenocor- tical carcinoma syndrome revisited. Am. J. Dis. Child., 139, 134-136 (1985).
9) Greenbert, R. S. and Schuster, J. E., Jr. Epidemiology of cancer in children. Epidemiol. Rev., 7, 22-48 (1985).
10) Birch, J. M. Manchester Children’s Tumor Registry 1954-1970 and 1971-1983. In “International Incidence of Childhood Cancer in the World,” ed. D. M. Parkin, C. A. Stiller, G. J. Draper, C. A. Bieber, B. Terracini and J. L. Young, IARC Scientific Publications No. 87, pp. 299-304 (1988). IARC, Lyon
11) “Japan Children’s Cancer Registry. Vol. 3 (1979-1983)” (1987). Children’s Cancer Association of Japan. Tokyo.
12) Research Group for Population-based Cancer Registration in Japan. Cancer incidence and incidence rates in Japan in 1985 - estimates based on data from seven population- based cancer registries. Jpn. J. Clin. Oncol., 20, 212-218
(1990).
13) Sondik, E. J. “Cancer Statistics Review 1973-1986, in- cluding a Report on the Status of Cancer Control May 1989,” NIH Publication No. 89-2789 (1989). US Depart- ment of Health and Human Services, Bethesda.
14) Breslow, N. and Day, N. E. “Statistical Methods in Cancer Research Vol. II. The Design and Analysis of Cohort Studies,” IARC Scientific Publications No. 82, pp. 69-71 (1987). IARC, Lyon.
15) Ise, T., Ohira, M., Omiya, A., Hasegawa, H. and Naka- jima, T. Primary adrenocortical carcinoma and ganglio- neuroblastoma followed by rhabdomyosarcoma in a child. A case report. Jpn. J. Clin. Oncol., 12, 397-404 (1982).
16) Li, F. Cancer families: human models of susceptibility to neoplasia - The Richard and Hilda Rosenthal Foundation Award Lecture. Cancer Res., 48, 5381-5386 (1989).
17) The Japanese Society of Pathology. “Annual of the Pathological Autopsy Cases in Japan (1974-1988),” Vol. 16-31 (1974-1988). The Japanese Society of Pathology, Tokyo.
18) Strong, L. C., Stine, M. and Norsted, T. L. Cancer in survivors of childhood soft tissue sarcoma and their rela- tives. J. Natl. Cancer Inst., 79, 1213-1220 (1987).
19) Birch, J. M., Hartley, A. L., Marsden, H. B., Harris, M. and Swindell, R. Excess risk of breast cancer in the mothers of children with soft tissue sarcomas. Br. J. Cancer, 49, 325-331 (1984).
20) Hartley, A. L., Birch, J. M., Marsden, H. B. and Harris, M. Breast cancer risk in mothers of children with osteosarcoma and chondrosarcoma. Br. J. Cancer, 54, 819-823 (1986).
21) Burki, N., Gencik, A., Torhorst, J. H., Weber, W. and Miller, H. Familial and histological analyses of 138 breast cancer patients. Breast Cancer Res. Treat., 10, 159-167 (1987).
22) Anderson, D. E. and Badzioch, M. D. Combined effects of family history and reproductive factors on breast cancer risk. Cancer, 63, 349-353 (1989).
23) Lynch, H. T., Albano, W. A., Layton, M. A., Kimberling, W. J. and Lynch, J. F. Breast cancer, genetics, and age at first pregnancy. J. Med. Genet., 21, 96-98 (1984).
24) Muir, C., Waterhouse, J., Smans, M., Casset, F. (ed.). “Cancer Incidence in Five Continents Vol V,” IARC Scientific Publications No. 88 (1987). IARC, Lyon.
25) Hartley, A. L., Birch, J. M., Kelsey, A. M., Marsden, H. B., Harris, M. and Teare, M. Are germ cell tumors part of the Li-Fraumeni cancer family syndrome? Cancer Genet. Cytogenet., 42, 221-226 (1989).
26) Little, J. B., Nove, J., Dahlberg, W. K., Troilo, P., Nichols, W. W. and Strong, L. C. Normal cytotoxic response of skin fibroblasts from patients with Li- Fraumeni familial cancer syndrome to DNA-damaging agents in vitro. Cancer Res., 47, 4229-4234 (1987).
27) Yano, T., Linehan, M., Anglard, P., Lerman, M. I., Daniel, L. N., Stein, C. A., Robertson, C. N., LaRoca, R. and Zbar, B. Genetic changes in human adrenocortical
carcinomas. J. Natl. Cancer Inst., 81, 518-523 (1989).
28) Yokota, J., Wada, M., Shimosato, Y., Terada, M. and Sugimura, T. Loss of heterozygosity on chromosomes 3, 13, and 17 in small cell carcinoma and on chromosome 3 in adenocarcinoma of the lung. Proc. Natl. Acad. Sci. USA, 84, 9252-9256 (1987).
29) Malkin, D., Li, F. P., Stong, L. C., Fraumeni, J. F., Jr., Nelson, C. E., Kim, D. H., Kassel, J., Gryka, M. A., Bischoff, F. Z., Tainsky, M. A. and Friend, S. H. Germ line p53 mutations in a familial syndrome of breast cancer, sarcomas, and other neoplasms. Science, 250, 1233-1238 (1990).