Important principles have been learned from studies of childhood cancer about the fundamental biology of neoplasia at all ages. The information thus derived gives new information about environmental exposures as well as host susceptibility.
Understanding etiology or mechanisms can lead to new methods for prevention, early detection, and therapy. An alert observation at the bedside often has been a first step in studies leading to such new understanding.
Environmental exposures can include transplacental carcinogenesis, lactational carcinogens, sunlight, and chemical, mineral, and viral contact during childhood.
Transplacental carcinogenesis. Seven years have elapsed since diethylstilbestrol (DES), the first known human transplacental carcinogen, was identified through the observation of a cluster of a rare neoplasm in young women that usually occurred only in older women.1* The risk of the cancer, clear-cell adenocarcinoma of the vagina or cervix, has since been defined as being between 0.14 and 1.4 cases per 1,000 exposed daughters.2 A much higher frequency of minor anomalies of the vagina and cervix occurred - adenosis, or ridging of the mucosa.3 Contrary to laboratory experimentation indicating that transplacental carcinogenesis entails intrauterine exposures after organogenesis,4 the DES-related cancers involved exposures before the 18th week of gestation - i.e., during organogenesis.2 Thus, the neoplasia and malformations appear to have origins in common.
Males were also affected by in utero exposures to DES - not with cancer as far as we know now, but with hypospermia and malformations of the genital tract, such as epididymal cysts.5 Other effects in males or females may lie ahead. One preliminary report suggests that perinatal morbidity may be high among the offspring of women who were themselves exposed in utero to DES. These unfavorable outcomes of pregnancy may be related to an abnormal shape of the uterus as seen on hysterosalpingography.6
No other transplacental human carcinogens have been identified, but one is strongly suspected. Diphenylhydantoin (DPH), taken for epilepsy, has been associated with lymphoma or, when used during pregnancy, can produce malformations in the child - typically, midfacial hypoplasia and aplasia of the finger- and toenails.* It was therefore thought that DPH taken during pregnancy might transplacentally induce lymphoma in the offspring, and it came as a surprise when two children with the fetal hydantoin syndrome were described not with lymphoma but with neuroblastoma (reviewed by Miller7).
The possibility that immunosuppressants given during pregnancy might induce lymphoma in the offspring has fortunately not been observed, even when given in conjunction with renal transplantation8,9 - a powerfully Iymphomagenic circumstance.10 Among other possible maternal exposures during pregnancy that may predispose the child to neoplasia are the use of alkylating agents for the treatment of disease11 or occupational exposure to benzene.12 In this connection, it is of interest that gas chromatographic- mass spectrometric analysis of low- molecular- weight volatile organic substances in umbilical cord blood and maternal blood samples revealed more than 100 components to be present in the cord blood in significantly higher concentrations than in the maternal blood. Notable among the chemicals in this study were benzene, carbon tetrachloride, and chloroform.13
Are there lactational carcinogens? Environmental pollutants can be found in breast milk. None is known to cause cancer in man, but one group - the polychlorinated biphenyls (PCBs) - is said to have retarded growth of breast-fed infants in Kyushu, Japan. In this incidence, which has been referred to by Dr. Arena elsewhere in this issue, mothers of newborns ate foods prepared in cooking oil that had been heavily contaminated with PCBs and several impurities.14
Polychlorinated biphenyls have also caused problems in the United States, where they have been dumped as industrial waste into waters used for sports fishing.15
Poiybrotninated biphenyls (PBBs), chemically similar to PCBs, contaminated much of Michigan when the chemical (used as a fire retardant) was confused with magnesium oxide^ a cattle-feed supplement.16 These chemicals, dioxin, and certain pesticides are deposited in fat, and best excreted in the fat of breast milk. The potential for harm to infants exists, but adverse effects, if any, are as yet unknown, except for the exposure of Japanese infants to breast milk containing PCBs as well as other impurities.
The Committee on Environmental Hazards of the American Academy of Pediatrics has issued a statement (see box) indicating that its members feel mothers should continue to be encouraged to breast-feed their infants unless there is a history of exposure to PCBs.
EXPOSURES DURING CHILDHOOD
Chemicals, sunlight, ionizing radiation, asbestos, and viruses are among the exposures a child or adolescent may encounter that may be related to in later life.
Chemicals. Anabolic androgenic steroid therapy for aplastic anemia, particularly of the Fanconi is associated with liver neoplasia, usually malignant.* There may be regression if the drug is discontinued in time.18,19 The anemia appears to susceptibility to the carcinogenic effects the drug.
Patients given immunosuppressive therapy for renal transplantation experience 150 times the usual frequency of reticulum cell sarcoma, particularly of the brain, with onset as soon as three months after treatment was initiated.10 The risk seems to be greatest in persons under 20 years of age. In one series, seven developed lymphomas as against 0.161 expected, and six developed other cancers, especially hepatobiliary, as against 0.649 expected.10
Alkylating agents used to treat cancer are associated with an excess of nonlymphocytic leukemia.11 These drugs are also used to treat such nonmalignant diseases as nephritis, rheumatoid arthritis, and scleroderma. In prescribing these drugs, the physician must keep in mind their possible neoplastic consequences.
Squamous cell carcinoma has occurred in the buccal mucosa of a 22-year-old electronics buff, who began at eight years of age to chew plasticcoated wire insulation and other types of plastic.20 He often stored the material in his mouth for up to eight hours.
Only a small fraction of human cancer can be related to specific chemical exposures other than cigarette smoking. Most cancers are currently thought to be induced by chemicals that form within the body by metabolic activation.21 Identification of these chemicals may be aided by applying bacterial mutagenicity tests or other screening procedures to urine, feces, or bile.
Laboratory studies suggest that a wide variety of chemicals in food may be carcinogenic. Such natural constituents as fat, anatoxin, or tannins may be involved, or substances may be introduced in processing or preparing food - e.g., polycyclic hydrocarbons from smoking or grilling meat or nitrites used as a preservative.21 Overnutrition is thought to contribute to the development of endometrial cancer,22 and fats are thought to influence the occurrence of cancer of the breast22 or colon.23
It has been difficult to derive evidence for carcinogenicity of individual components of the diet because of the long latent periods, metabolic activation, and possible interactions with host susceptibility and other agents, the effects of which may be enhancing or inhibiting.
Sunlight. Overexposure of children to sunlight may contribute to the occurrence of skin cancer later in life, but when the genetic disorder xeroderma pigmentosum (XP) is present, the process is speeded up, so the skin neoplasia often occurs during childhood. In North Africa, XP is unusually prevalent, and at the Institut Salah Aziaz in Tunis, 14 percent of all childhood cancer is squamous cell carcinoma on XP.24
Ionizing radiation. The carcinogenic effects of gamma radiation are well established from studies of Japanese atomic-bomb survivors,25 and are being increasingly seen, as radiotherapy for cancer extends survival.24,26 The capacity of intrauterine exposures to induce each form of childhood cancer equally, as reported by Stewart and Kneale,27 has not been found in recent studies.28,28 The most dramatic new development concerning radiation exposures has been the discovery of an interaction with certain rare genetic diseases characterized by defective repair of DNA and a propensity to cancer (see below).
Asbestos. Although asbestos does not cause cancer during childhood, exposures during childhood cause mesothelioma decades later.30 In one dramatic example, a man with asbestosis and lung cancer had worked at a shipyard as a pipe insulator for 25 years. He wore his dusty clothes home, where his family was exposed to them. Mesothelioma developed in his wife at 50 years of age and in his daughter at 34.31 Although asbestos is a chemical agent, it apparently is carcinogenic because of the physical properties of its fibers (thin diameter and length of 8 μ or greater).32
Viruses. In the 1960s, virologists hoped soon to show on the basis of their experimentation on mice and cats that human leukemia was virally induced. Various epidemiologic tests of their hypotheses failed to show a human counterpart for their experimental observations. Great attention focused on time-space clusters of childhood leukemia, but these were never shown to be horizontally transmitted in a fashion that would suggest infectious origins.33 Two new developments, however, indicate that under special circumstances at least, Epstein-Barr virus (EBV) is oncogenic - in Burkitt's lymphoma in Africa34 and in the X-linked lymphoproliferative disorder described by Purtilo and his associates.35 The last-mentioned disorder affects several boys in families with a lymphomalike disease, agammaglobulinemia, or fatal infectious mononucleosis - each of which is a different manifestation of B-lymphocyte disease.
Nasopharyngeal carcinoma is also likely to be due in some geographic areas to EBV infection.36 Carcinoma of the uterine cervix has epidemiologic characteristics that strongly suggest that it is due to a transmissible agent, and it has indeed been described as a venereal disease.37 Sexual activity earlier in life now brings the origins of this neoplasm down into the pediatric age range.
Other evidence. Although a relatively small proportion of all cancer is attributable to environmental factors other than cigarette smoking and sunlight, as yet unidentified environmental factors are indicated by the change in cancer rates when people migrate from one country to another,38 by maps of cancer mortality by county in the United States, 1950-69, which show distinctive patterns of excessive mortality from specific forms of cancer,39 and by differences in life style, as among Seventh-Day Adventists, whose total cancer mortality rates are 50 to 70 per cent of those for the general population for most sites unrelated to the carcinogenic effects of smoking or drinking.40 In the aggregate, these studies indicate that exogenous agents encountered early in life can influence cancer rates for many years to come.
There is a near absence of specific neoplasms among certain ethnic groups. For example, blacks living in the United States rarely develop Ewing's tumor. There is a peak in the occurrence of acute lymphocytic leukemia among white children living in the United States at the age of four, but blacks have yet to display such a peak.24
Studies of increased susceptibility of persons to specific cancers have been more informative biologically than have studies of inherent resistance to cancer. For example, three main childhood cancers - retinoblastoma, Wilms' tumor, and neuroblastoma - have been epidemiologically separated into heritable and sporadic types.41 Each is thought to require two events. In the heritable form, usually multifocal, the first event is a germinal mutation that affects all target cells, sometimes in more than one organ. The second event is thought to be a somatic mutation, sufficiently common for the tumor to be bilateral or otherwise multifocal. In the sporadic form, both events are thought to be postzygotic - i.e., a double somatic mutation, unlikely to occur by chance in the same cell, accounting for the unifocal nature of the neoplasm.
A more extended consideration of the genetics of childhood cancer is beyond the scope of this presentation. Interested readers may find further details elsewhere.42 Of particular interest is the dramatic interaction of cancer-prone genetic disorders and radiant energy.
When children with ataxia-telangiectasia, an autosomal-recessive trait with a marked predisposition to lymphoma,43 were treated with conventional doses of radiotherapy, they sometimes suffered severe acute radiation reactions.44 Exposure of their skin fibroblasts in culture to various doses of x-ray caused diminished survival in proportion to the dose45 and revealed a DNA-repair defect.46 At least two complementation groups and one variant have been identified. The defect in ataxia-telangiectasia is thus an x-ray analogue of the DNA-repair defect after ultraviolet light damage in children with xeroderma pigmentosum.
Cells from patients with these two diseases are being used to screen for chemical mutagens/ carcinogens. Chemicals do not cause DNA-repair defects in XP, but certain carcinogens increase the frequency of sister-chromatid exchanges.47 A recent report from Sweden indicates that studies of sister-chromatid exchanges in the newborn may be informative with regard to maternal exposures during pregnancy - e.g., to benzene.12
These studies represent but a small fraction of those in pediatrics that have made immense contributions to understanding of the biology of cancer.
THE NEED FOR CLINICAL ETIOLOGISTS
Much has been learned about carcinogenesis in recent years from studies of cancer in children or from studies of exposures dating back to childhood or even intrauterine life. New insights often begin with alert observations at the bedside. At present, however, there is a dearth of bedside etiologists. Ample numbers of pediatric oncologists are available for studies relating to diagnosis and treatment - but not to etiology.
Consideration of cause, when it is said to be unknown, does not come easily to many physicians. Such thinking requires a quirk of mind not often found in medicine (although not at all uncommon in other fields, where it is a part of everyday life - e.g., politics, business, journalism, and police work).
It may be helpful to distinguish between traditional epidemiologists and clinical etiologists. To me, the traditional epidemiologist is strong in his knowledge of scientific design but less so in his knowledge of clinical and laboratory medicine. The traditional epidemiologist is a master in testing hypotheses, especially when detecting a small effect vs. no effect is difficult. The clinical etiologist has a natural curiosity about cause as it relates to the specialist's comprehension of disease and tends to generate hypotheses based on clinical observations and recent developments in laboratory research.
The ranks of recruits to clinical etiology are thin and are unlikely to be increased by increasing funds for training programs. That was tried in the 1950s with very little success.
What is needed at cancer centers and at universities is a spirit of inquiry about disease causation. When the patient is first seen, a serious attempt should be made to determine if the personal or family history or physical findings are peculiar. It is peculiarities of occurrence that can provide new understanding about the origins of disease. Before treatment is started, the physician should consider what specimens should be obtained before findings may be confused by the effects of drugs. One must decide what specimens should be obtained for immediate study or be stored (e.g., skin fibroblasts) for study when new techniques are developed.
Diagnosticians and therapists are unlikely to be wooed to this additional clinical interest. A more promising approach would be to develop a clinical etiologist for each cancer center, who would make rounds with the staff regularly and concentrate on enologie findings in a search for exceptional cases that provide special research opportunities.
1. Herbst, A. L., Ulfelder, H., and Poskanzer, D. C. Adenocarcinoma of the vagina. N. Engl. J. Med. 284 (1971), 878-881.
2. Herbst, A. L., Cole, P., Colton, T., Robboy, S. J., and Scully, R. E. Age-incidence and risk of diethylstilbestrol-related clear cell adenocarcinoma of the vagina and cervix. Am. J. Obstet. Gynecol. 128 (1977), 43-50.
3. Robboy, S. )., Scully, R. E., Welch, W. R., and Herbst, A. L. Intrauterine diethylstilbestrol exposure and its consequences: Pathologic characteristics of vaginal adenosis, clear cell adenocarcinoma, and related lesions. Arch. Pathol. Lab. Med. 101 (1977), 1-5.
4. Rice, J. M. An overview of transplacental chemical carcinogenesis. Teratology 8 (1973), 113-126.
5. Bibbo, M., et al. Follow-up study of male and female offspring of DES-exposed mothers. Obstet. Gynecol. 49 (1977), 1-8.
6. Kaufman, R. H., Binder, G. L., Gray, P. M., Jr., and Adam, E. Upper genital tract changes associated with exposure in utero to diethylstilbestrol. Am. J. Obstet. Gynecol. 128 (1977), 51-59.
7. Miller, R. W. Environmental causes of cancer in childhood. In Bamess, L. A. (ed.). Advances in Pediatrics, Volume 25. Chicago: Year Book Medical Publishers (in press).
8. Cederqvist, L. L., Merkatz, LR., and Litwin, S. D. Fetal immunoglobulin synthesis following maternal immunosuppression. Am. J. Obstet. Gynecol. 129 (1977), 687-690.
9. Papoff, P., Whetham, J. C, Katz, A., and deVeber, G. A. Pregnancy in renal transplant recipients: report of two successful pregnancies in a patient with impaired renal function. Can. Med. Assoc. J. 117 (1977), 1288-1295.
10. Hoover, R. Effects of drugs - immunosuppression. In Hiatt, H. H., Watson, ). D., and Winsten, J. A. (eds.). Origins of Human Cancer. A. Incidence of Cancer in Humans, Volume 4. Cold Spring, N.Y.: Cold Spring Harbor Laboratory, 1977, pp. 369-379.
11. Reimer, R. R., Hoover, R., Fraumeni, J. F., Jr., and Young, R. C. Acute leukemia after alkylating-agent therapy of ovarian cancer. N. Engl. J. Med. 297 (1977), 177-181.
12. Funes-Cravioto, F., et al. Chromosome aberrations and sister-chromatid exchange in workers in chemical laboratories and a rotoprintíng factory and in children of women laboratory workers. Lancet 2 (1977), 322-325.
13. Dowry, B. J., Laseter, J. L., and Storer, J. The transplacental migration and accumulation in blood of volatile organic constituents. Pediatr. Res. 10 (1976), 696-701.
14. Kuratsune, M., Yoshimura, T., Matsuzaka, J., and Yamaguchi, A. Epidemiologic study on Yusho, a poisoning caused by ingestion of rice oil contaminated with a commercial brand of polychlorinated biphenyls. Environ. Health Perspect. 1 (1972), 119-128.
15. Graham, J. M. Levels of PCB's in Canadian commercial fish species (abstr.). In Proceedings of the National Conference on Polychlorinated Biphenyls. Washington, D.C.: Environmental Protection Agency, Office of Toxic Substances, EPA -560/6-75-004, 1976, pp. 155-160.
16. Stadtfeld, C. G. Cheap chemicals and dumb luck. Next time they may be burying people instead of cows. Audubon (January, 1976), 110-118.
17. Committee on Environmental Hazards, American Academy of Pediatrics. American Academy of Pediatrics News and Comments (April, 1978), 5.
18. Johnson, F. L., et al. Association of androgenic-anabolic steroid therapy with development of hepatocellular carcinoma. Lancet 2 (1972), 1273-1276.
19. Meadows, A. T., Naiman, J. L., and VaJdes-Dapena, M. Hepatoma associated with androgen therapy for aplastic anemia. /. Pediatr. 84 (1974), 109-110.
20. Casterline, C. L., Casterline, P. F., and Jaques, D. A. Squamous cell carcinoma of the buccal mucosa associated with chronic oral polyvinyl chloride exposure: report of a case. Cancer 39 (1977), 1686-1688.
21. Miller, E. C. Some current perspectives on chemical carcinogenesis in humans and experimental animals. Presidential address. Cancer Res. 38 (1978), 1479-1496.
22. Cole, P., and Cramer, D. Diet and cancer of endocrine target organs. Cancer 40 (1977), 434-437.
23. Reddy, B. S., Mastromariono, A., and Wynder, E. Diet and metabolism: large-bowel cancer. Cancer 39 (1977), 1815-1819.
24. Miller, R. W. Cancer and congenital malformations: another view. In Mulvihill, J. H., Miller, R. W., and Fraumeni, J. F., Jr. (eds.). Genetics of Human Cancer. New York: Raven Press, 1977, pp. 77-79.
25. Miller, R. W. (ed.). Cancer research by the Atomic Bomb Casualty Commission. A compilation of editorials published in the Journal of the National Cancer Institute. Natl. Cancer Inst. Monogr. 52 (in press).
26. Meadows, A. T., et al. Patterns of second malignant neoplasms in children. Cancer 40 (1977), 1903-1911.
27. Stewart, A., and Kneale, G. W. Changes in the cancer risk associated with obstetric radiography. Lancet 1 (1968), 104-107.
28. Hutchison, G. B. Late neoplastic changes following medical irradiation. Cancer 37 (1976), 1102-1107.
29. Robinette, C. D., and Jablon, S. Childhood cancer and fetal x-ray exposure in children born in military hospitals (abstr.). Radiat. Res. 67 (1976), 627.
30. Anderson, H. A., et al. Household-contact asbestos. Neoplastic risk. Ann. N. Y. Acad. Sa. 271 (1976), 311-323.
31. Li, F. P., et al. Familial mesothelioma after intense asbestos exposure at home. J. A.M. A. 240 (1978), 467.
32. Stanton, M. F., et al. Carcinogenicity of fibrous glass: pleural response in the rat in relation to fiber dimension. J. Natl. Cancer Inst. 58 (1977), 587-603.
33. Caldwell, G. G., and Heath, C. W., Jr. Case clustering in cancer. South. Med. J. 69 (1976), 1598-1602.
34. de-Thé, G., et al. Epidemiological evidence for causal relationship between Epstein-Barr virus and Burkitt's lymphoma from Ugandan prospective study. Nature 274 (1978), 756-761.
35. Purtilo, D. T., et al. Variable phenotypic expression of an X-linked recessive lymphoproliferative syndrome. N. Engl. I. Med. 297 (1977), 1077-1081.
36. Rapp, F., and Reed, C. L. The viral etiology of cancer: a realistic approach. Cancer 40 (1977), 419-429.
37. Kessler, I. I. Human cervical cancer as a venereal disease. Cancer Res. 36 (1976), 783-791.
38. Haenszel, W. Migrant studies. In Fraumeni, J. F., Jr. (ed.). Persons at High Risk of Cancer: an Approach to Cancer Etiology and Control. New York: Academic Press, 1975, pp. 361-371.
39. Blot, W. J., Fraumeni, J. F., Jr., Mason, T. J., and Hoover, R. Developing clues to environmental cancer, a stepwise approach using cancer mortality data. Environ. Health Perspect. (in press).
40. Phillips, R. L. Role of life-style and dietary habits in risk of cancer among Seventh-Day Adventists. Cancer Res. 35 (1975), 35133522.
41. Knudson, A. G., Jr. Mutation and human cancer. Adv. Cancer Res. 17 (1973), 317-352.
42. Mulvihill, J. J., Miller, R. W., and Fraumeni, J. F., Jr. (eds.) Genetics of Human Cancer. New York: Raven Press, 1977.
43. Kersey, J. H., Spector, B. D., and Good, R. A. Primary immunodeficiency diseases and cancer: the Immunodeficiency-Cancer Registry. Int. J. Cancer 12 (1973), 333-347.
44. Cunliffe, P. N., et al. Rathosensitivity in ataxia- telangiectasia. Br. J. Radiol. 48 (1975), 374-376.
45. Taylor, A. M. R., Metcalfe, J. A., Oxford, J. M., and Harnden, D. G. Is chromatid-type damage in ataxia telangiectasia after irradiation at GD a consequence of defective repair? Nature 260 (1976), 441443.
46. Paterson, M. C, et al. Defective excision repair of y-raydamaged DNA in human (ataxia telangiectasia) fibroblasts. Nature 260 (1976), 444-447.
47. Wolff, S., Rodin, B., and Cleaver, J. E. Sister chromatid exchanges induced by mutagenic carcinogens in normal and xeroderma pigmentosum cells. Nature 265 (1977), 347-349.