Cancer Genesis Across the Age Spectrum: Associations With Tissue Development, Maintenance, and Senescence

Cancer Genesis Across the Age Spectrum: Associations With Tissue Development, Maintenance, and Senescence

Cancer Genesis Across the Age Spectrum: Associations With Tissue Development, Maintenance, and Senescence Philip Rubin, MD,* Jacqueline P. Williams, P...

1MB Sizes 1 Downloads 83 Views

Cancer Genesis Across the Age Spectrum: Associations With Tissue Development, Maintenance, and Senescence Philip Rubin, MD,* Jacqueline P. Williams, PhD,* Susan S. Devesa, PhD,† Lois B. Travis, MD, ScD,* and Louis S. Constine, MD*,‡ Cancer genesis across the age spectrum is a complex, multifactorial process, and parallels changes in site-specific tissue development, maintenance, and senescence. Cancer is not a single disease, and different tumor and stem cells may demonstrate various manifestations of abnormal function. Mutations in DNA, some random and some explained by exogenous insults, accompanied by changes in the tissue microenvironment, generally precede the onset of aberrant replication and apoptosis. Moreover, increasing evidence suggests that genetic programs normally active only during development of human beings may be reactivated during tumorigenesis. The complicated underlying biology of human growth, development, and carcinogenesis is reflected in the highly disparate patterns in site-specific cancer incidence rates across age groups. In childhood, the peak years of an organ system’s increase in size correlate with peak years of cancer incidence. Conversely, in most adult-onset cancers, it is exposure to exogenous toxins, the failure of maintenance and repair, and finally, dysfunction(s) in the normal cellular aging process that likely play a role in the development of these malignancies. Additional basic science investigations and epidemiologic studies will assist in our understanding of the mechanisms that underlie the notable difference in site-specific cancer incidence according to age. Semin Radiat Oncol 20:3-11 © 2010 Elsevier Inc. All rights reserved.

C

ell division and differentiation are the dominant features of pre- and postnatal development in all mammals, including human beings. These functions continue during the “homeostatic” period of adulthood, although cellular repair begins to play a greater role during what may be considered a “maintenance phase.” However, as tissues and organs age further, they undergo progressive degenerative changes and gradually lose their proliferative homeostasis because of many mechanisms, including apoptosis and senescence.1 By contrast, tumorigenesis involves the successful escape of cells from the boundaries that impede unrestrained division and the acquisition of adaptive mechanisms that permit survival, both in the tissue of origin and throughout the body. Our understanding of the mechanisms by which this phenomenon occurs has increased exponentially

*Department of Radiation Oncology, University of Rochester Medical Center, Rochester, NY. †Division of Cancer Epidemiology and Genetics, National Cancer Institute, NIH Department of Health and Human Services, Bethesda, MD. ‡Department of Pediatrics, University of Rochester Medical Center, Rochester, NY. Address reprint requests to Louis S. Constine, MD, Department of Radiation Oncology, James P. Wilmot Cancer Center, University of Rochester Medical Center, 601 Elmwood Avenue, Box 647, Rochester, NY 14642. E-mail: [email protected]

1053-4296/10/$-see front matter © 2010 Elsevier Inc. All rights reserved. doi:10.1016/j.semradonc.2009.08.001

over the past 2 decades, especially with the expansion of genomic and proteomic techniques. Despite these advances, an explanation for why different cancers develop in disparate tissues at varying developmental stages remains largely speculative, although some plausible or determinant etiologies are identifiable. Using molecular, genetic, and epidemiologic data to assist us in dissecting some of the inherent complexities in this arena, we will explore the relationship of human age, development, attempt to maintain homeostasis, and the failure of these mechanisms on the observed distribution of cancers across the age spectrum.

The Biology of Cancer Genesis Epidemiologic data demonstrate that the patterns of oncogenesis in different tissues and organs vary across the human lifespan, with specific tumors being associated with different phases in life. This variation can be observed in terms of both tumor type and their response to treatment. With our increased understanding of the biology of normal stem cells and the more recent recognition of the so-called “cancer stem cell,”2,3 it is intriguing to speculate that the differential development of neoplasias over the lifespan could, in part, relate to genetic mutations that alter the progressive and changing role that stem cells play over time 3

P. Rubin et al

4

Age-specific Incidence Rates for Various Cancers, SEER- 9 Registries, 1973-2005.

A

Epithelial

Hepatic and Renal

Sex Specific

1000

1000

100

100

100

10

10

10

1

1

1

0.1

0.1

0.1

Rate per 100,000 person-years

1000

0.01

0.01

0.01 0

10

20

30

40

50

60

70

80

Breast (Female only) Female Genital System Prostate Testis

Hepatic Renal

Oral Cavity and Pharynx Alimentary Tract Lung and Bronchus

90 100

0

10

Lymphomas and Leukemias

20

30

40

50

60

70

80

0

90 100

100

100

100

10

10

10

1

1

1

0.1

0.1

Rate per 100,000 person-years

1000

0.1

Hodgkin Lymphoma Non-Hodgkin Lymphoma Acute Lymphocytic Leukemia Acute Myeloid Leukemia Chronic Leukemia 0.01 30

40

50

60

70

80

90 100

40

50

60

70

80

90 100

Fibrosarcoma Osteosarcoma Rhabdomyosarcoma

Glial Embryonal 0.01

0.01 20

30

Sarcomas

1000

10

20

Central Nervous System

1000

0

10

0

10

20

30

40

50

60

70

80

90 100

0

10

20

30

40

50

60

70

80

90 100

-----------------------------------------------------------------------------------------------Age at diagnosis-----------------------------------------------------------------------------------------------

Figure 1 Age-specific incidence rates for various cancers, Surveillance, Epidemiology, and End Results Program, 9 registries, 1973-2005.15-17 (Color version of figure is available online.)

Cancer genesis across the age spectrum

5

Age-specific Incidence Rates for Various Pediatric and Young Adult Cancers, SEER- 9 Registries, 1973-2005.

B

Rate per 100,000 person-years

Lymphomas and Leukemias

Central Nervous System

Hepatic, Renal and Testis

Sarcomas

20

20

20

20

10

10

10

10

1

1

1

1

0.1

0.1

0.1

0.1

Hodgkin Lymphoma Non-Hodgkin Lymphoma Acute Lymphoblastic Leukemia

Glial Embryonal

0.01

0.01 0

5

10

15

20

25

30

Osteosarcoma Rhabdomyosarcoma

0.01

0.01 0

5

10

15

20

25

30

Testis Hepatic Renal

0

5

10

15

20

25

30

0

5

10

15

20

25

30

---------------------------------------------------------------------------Age at diagnosis---------------------------------------------------------------------------

Figure 1 (continued).

if, as proposed by some, the cancer stem cells arise from the normal stem cell pool. However, if cancer is a stochastic event and can arise from any cell progeny, then the tumorigenic factors leading to such induction expand beyond the simple premise of genetic mutation within the stem cell population, and will now include environmental stressors and alterations in phenotypic responses in the somatic as well as the stem cell populations.4 From the hypothesis that cancer stem cells originate from normal stem cells, mutation(s) in the most basic and earliest stem cell, the pluripotent embryonic stem cell, would result in far-reaching consequences because of the criticality of this developmental stage. Indeed, cytotoxic insult within the first month of embryogenesis after implantation usually leads to prenatal death and spontaneous abortion.5 It also can be visualized that more subtle mutations (ie, such less lethal changes) could have more wide-ranging outcomes, such as teratogenesis (or development of other congenital defects) or even carcinogenesis,6 with the latter leading to the induction of embryonal tumors. An association between congenital defects and neoplasia is well documented; an excellent illustration is Wilms tumor, which is associated with congenital birth defects, such as aniridia, hemihypertrophy of limb, genitourinary malformations, and mental retardation. Such a link between oncogenesis and early developmental defects has been recognized for over a century. In the late

1800s, for example, 2 independent investigators suggested that embryonal tumors arose from developmental defects within excess embryonic tissues,7,8 cells that are normally removed through developmental apoptotic pathways.9,10 Of course, since the 1800s, we have gained a greater understanding of the role of the cellular milieu in cell function and are aware that embryonic stem cells are sustained within a microenvironment that facilitates self-renewal, and that many cancer cells release and respond to a convergence of similar growth signals, such as the Notch, Wingless, and transforming growth factor ␤ superfamilies. However, as was described in the now seminal paper from Hanahan and Weinberg,11 there are 6 hallmarks of cancer that encompass the definition of, and various roles played by, oncogenes and tumor suppressor genes. These hallmarks include not only self-sufficiency in growth signaling and limitless replicative potential, but also insensitivity to antigrowth signals and evasion of apoptosis. The embryonic stem cell environment not only facilitates self-renewal, but also maintains a balance between renewal and differentiation. Interestingly, some investigators have demonstrated that transfer of transformed cancer cells into an embryonic stem cell microenvironment will cause an imposition of the same regulatory controls on cancer cell growth,12 confirming the rigors of cell growth regulation. Many tumor types, such as breast and colorectal, seem to result from the cellular escape of

P. Rubin et al

6

Figure 2 Diagrammatic representation depicting the proposed mechanisms of tumor initiation and progression across the lifespan of human beings. (Modified from Vet Pathol 46:176-193, 2009, and Aging Res Rev 8:94-112, 2009.)

regulatory control through the inappropriate upregulation of developmental pathways, such as Wnt and Notch.13,14 After human development moves into the postnatal period, proliferation (ie, growth within the progenitor cell compartments, rather than the relatively m tumorigenesis ore quiescent stem cell compartment per se) is likely to provide the more likely milieu for tumorigenesis. It is during “spurts” in the rapidly growing tissues and organs that those mutations would be readily amplified, particularly during the adolescent phase. If indeed there is such a dependence on cell division to provide the environment for tumor induction, then reaching adulthood (and growth cessation) should result in a plateau, and possibly a decline in tumor incidence. Such a pattern can be readily identified in osteosarcomas. However, this pattern is almost nonexistent in other tumor types and, instead, most tumor incidence rates indicate a relatively steep increase in incidence during the second to sixth decades of life when many tissues and organs are homeostatic with respect to growth (Fig. 1A,B). Therefore, during this life phase, we speculate that the greatest factors driving oncogenesis are the accumulation of toxic and environmental insults, which would be expressed more in those tissues that maintain the largest proliferative compartment (ie, epithelial and endothelial). A diagrammatic scheme depicting the possible mechanisms of carcinogenesis is presented in Figure 2. However, such an accumulation of damage would not account for the changes in tumor patterns during old age when repair mechanisms are likely to be less efficient, so that tumor incidence should rise and not plateau or even fall, as illustrated for several cancers in Figure 1A. Other possible explanations for the plateau or decrease in tumor incidence in the very elderly include (1) an age-related loss of proliferative homeostasis and the increasing numbers of cells that enter a senescent state, as described earlier in the text, and (2) underdiagnosis of

cancers in an aging population (eg, due perhaps to less aggressive diagnostic testing or other influences). Additional factors may also be reflected in the patterns observed for several cancer sites associated with tobacco use. For example, although the 30⫹ year cross-sectional age-specific curve for lung and bronchus cancer peaks at ages 75-79, the cohort-specific (age-specific) curves continued to increase throughout all ages until recent years; however declines seen recently at older ages probably reflect the decrease in smoking and possibly lower tar cigarettes. Similarly, the 30⫹ year crosssectional curve for oral cavity and pharyngeal peaks at ages 7074; the cohort-specific patterns are quite similar to those for lung and bronchus, with rising rates across all ages until recently, likely mirroring the changing frequency of cigarette smoking, but with an added effect of declines in hard liquor consumption. The changing slope for cancers of the female genital system taken together probably reflects the changing relative contributions of the component cancers, for example, uterine cervix (relatively young age at onset), uterine corpus, and ovary. The clear perimenopausal change in slope for breast cancer incidence rates most likely reflects changes in hormonal milieu.

Human Development and Cancer Genesis Using classic concepts of development, we recognize that the transition from prenatal to postnatal organogenesis and tissue maturation reflects the transition of cellular proliferation from pluripotent to differentiated cells. In this article, we refer to this process as “neogenesis.” Pluripotent cells are capable of evolving into a variety of cell types, but are unable to implant because

Cancer genesis across the age spectrum they are programmed to follow select pathways and respond to signals from other cells. As tissues form, mutations can result in congenital defects or rapidly amplify and lead to neoplasia. As previously noted, the association of congenital defects and neoplasia is well documented. The postulate is that many pediatric malignancies are inherently different from adult cancers because of the dual nature of embryonic growth. The embryonal tumors observed in children may be considered the outcome of the plasticity found within the developmental environment, which normally leads to neogenesis, but instead provides the promotional impetus for neoplasia. If oncogenesis is to be related to development beyond the prenatal period, we must first define the relevant developmental periods (Fig. 3). ● ●



Infancy/early childhood is defined as age 1-6 years, during which most tissues are rapidly developing. Late childhood is defined as age 6 years to puberty, when there is greater heterogeneity of tissue development; some tissues are dominated by cell proliferation, others by hypertrophy, and others are reasonably quiescent until the hormonal milieu changes at puberty. Puberty is defined as those years, leading into adulthood, during which an acceleration in development of several tissues occurs in response to the hormonal surges during these years.

Interestingly, the 5 major normal tissues and/or organ and/or system each have different growth spurts.

7 1. Blood cell formation and bone marrow (which assumes this role at 6 months in utero) continue to rapidly develop at birth, peaking between 5 and 10 years of age, but then involuting in the peripheral bones throughout life. 2. Lymphoid tissue is also expanding rapidly at multiple sites (eg, thymus, spleen, lymph nodes, mucosal loci in intestinal Peyer’s patches, appendix, as well as bone marrow), peaking between 6 and 12 years of age. 3. Central nervous system (CNS) brain neocortex continues its prenatal development with the most active phases of synaptogenesis and myelinization in the 1-5 years of age, but continuing for the subsequent 2-3 decades without significant volumetric change. Other organs with similar developmental dynamics include the kidney, heart, lungs, and liver. 4. Musculoskeletal growth is bimodal with the first 5 years of rapid growth followed by a second growth spurt at puberty. Regenerative potentials of muscle are heterogenous: cardiac myocytes have virtually none (cell growth is by hypertrophy, not replication, explaining the rapid postnatal development), skeletal muscle has limited capacity, whereas smooth muscle is highly regenerative. Other tissues with a bimodal pattern include gastrointestinal, head and neck (H&N), skin, and the vasculature. 5. Gonadal germ cells are relatively quiescent at birth and through the first 10 years of life when hormonal activation leads to sexual maturation between 12 and 18 years of age. Thus, each phase of rapid physical development in a normal tissues and/or organ and/or system coincides with the peak years of its derivative malignancies. This will now be discussed in terms of relevant cancer incidence rates.

Blood and Bone Marrow Neogenesis Blood and bone marrow comprise large tissue mass in terms of both volume and weight. Blood and bone marrow is a specialized connective tissue that consists of suspended hematopoietic cells in plasma (5-6 L). It accounts for 4%-5% body weight (1.6-3.7 kg), with a total weight exceeding that of liver. During fetal life and until 2 weeks after birth, hematopoiesis is in log phase growth; blood cells are also forming in liver and spleen which is transposed to bone marrow in children. Bone marrow extends to the full-length of long bones in addition to vertebrae and flat bones of pelvis, skull, ribs, and sternum. The bone marrow gradually involutes to the proximal femur and humerus, vertebrae, and flat bones (as noted earlier in the text) as the child grows into an adult (Fig. 4).

Figure 3 Growth curves of different tissues. (Modified from Tanner JM: Growth at Adolescence (ed 2). Oxford, UK, Blackwell Scientific Publications, 1962, pp 11.)

Neoplasia Neoplasia in the form of leukemia is the most common malignancy in children. The most frequent type is acute lymphoblastic leukemia (ALL) accounting for 80% of childhood leukemia, and acute myeloblastic leukemia occurring in another 20%. True mixed lineage leukemia express both lymphoid- and myeloid-associated antigens. Advances in our understanding of ALL have identified specific cytogenetic abnormalities in 60% of all cell lines. It is the rapid expansion of the bone marrow that may predispose the patient to develop overt leukemia. Thus, a

P. Rubin et al

8

Brain and Nervous Tissue

Figure 4 The relative amount of bone marrow in different anatomic sites as a function of age. (Reprinted with permission from Int J Radiat Oncol Biol Phys 31:1319-1339, 1995.)

sharp peak in ALL is observed in 2-3 year olds (⬎ 80 per million) that decreases to a rate of 20 per million for 8-10- year olds. It is estimated that leukemia comprises 17% of all cancers in the first year of life, increasing to 46% for 2 and 3 year olds and decreasing to 9% for 9 year olds, three-fourths of which are the ALL type (Fig. 1B).15

Lymphoid Tissues Neogenesis Lymphoid tissues protect the body by establishing the immune system and comprise numerous sites, such as regional lymph nodes, thymus, spleen, liver, extranodal sites like tonsils, Peyer’s patches in gut, mucosal associated tissue (MALT) in respiratory, genitourinary and gastrointestinal subepithelial loci, and ubiquitous sites throughout bone marrow. The lymphoid tissues rapidly increase in the first decade of life, peak at 10-14 years, and then involute over the second decade. Specific regions of each node are seeded with B lymphocytes from the bone marrow and T lymphocytes from the thymus. Neoplasia Lymphomas, which constitute about 15% of cancer diagnoses in patients aged ⬍ 20 years, are the third most common form of childhood cancer. They peak in incidence from age 7-11 years, similar to the volumetric expansion of the lymphoid system (Fig. 1B). Because absolute numbers are small, their effect on the overall incidence curves is moderate. Congenital immunodeficiency syndromes and acquired immunodeficiency syndrome are associated with an increased risk of non–Hodgkin lymphoma, though the contribution of these genetic diseases to the overall frequency is even smaller. Pediatric non–Hodgkin lymphoma is mostly high grade and includes major subtypes: precursor B and precursor T-lymphoblastic lymphoma, anaplastic large cell, and Burkitt lymphoma. Hodgkin’s lymphoma (HL) accounts for 6% of childhood malignancies with a male predominance ratio of 4:1. It is more common in children aged ⬎ 5 years and in adolescents. Incidence rates of HL then decline until the early adult years, when rates resurge. A relationship of HL to infection and early immunologic modulation (also from infectious exposures) may dominate as explanations for the incidence patterns of this malignancy.

Neogenesis Development of the brain is extremely rapid, both prenatally and in the early postnatal years. The brain’s prenatal morphology at 3 months includes 1 mm cerebral hemispheric cortex that enlarges to 8.0 mm at 6 months, and to 10.5 mm at 9 months (birth). Cortical enlargement is an evolutionary novelty related to higher mental function in mammals and is referred to as neocortex. The human brain at birth is 25% of its adult weight, doubles to 50% at 6 months, 75% at 2.5 years, 90% at 5 years, and 95% at 10 years. Elaboration of the frontal and parietal lobes volumetrically exceeds the occipital and temporal lobes and are the preferential sites for tumefaction. More than 100 billion neurons (gray matter) are estimated to comprise the nervous system, and these are embryonally derived and essentially are postmitotic cells compared with the glial cell population, including astrocytes (white matter). These glial cells derive from neuroepithelium and differentiate into several functionally diverse cells. Astrocytes play a supportive role and provide astrocytic end-feet that cover capillaries and create the blood brain barrier. Oligodendrocytes are also neuroepithelial, are smaller than astrocytes, and produce myelin sheaths. Ependymal cells are also thought to be neuroepithelial in origin and line brain ventricles. We now recognize the existence of subependymal cells that are believed to be stem cells capable of replication. Neoplasia It is estimated that 15%-20% of neoplasms in children arise in the CNS. The most common tumors reflect the earlier identified cell types, numbers, and locations. ● ● ●

Astrocytomas are mainly supratentorial Medulloblastomas arise predominantly in the cerebellum Ependymomas are largely mainly infratentorial, although they may occur in any location

The concept that neoplasia is more common in young children when an organ is exhibiting rapid growth is likely applicable to brain tumors. The highest incidence of CNS tumors in children occurs from infancy to 7 years of age and then decreases between the ages of 7 and 17. As shown in Figure 1B, the embryonally derived tumors (including medulloblastoma and primitive neuroectodermal [PNET] tumors) are most frequent in the first years of life and become progressively less common throughout the remainder of the human lifespan. These tumors are most likely consequences of genetic insults occurring prenatally, and thus coincide with the most rapid development of these cells. Conversely, the neuroepithelial tumors occur throughout childhood, adolescence, and adulthood, reflecting the continued activities of the associated cell populations. Specifically, astrocytes constitute the most proliferative group of cells, and hence astrocytomas are the most common CNS neoplasms throughout childhood. Finally, germ cell tumors (GCT) and teratomas, common during the pubertal and adolescent years, usually occur in diencephalic structures and seem almost exclusively as third ventricular lesions. Several genetic syndromes increase the propensity for CNS tumor

Cancer genesis across the age spectrum

9

development, such as neurofibromatosis. Although primary brain tumors constitute about 20% of all cancers in children, only 1% of cancers are primary brain tumors in adults because most cancers in this age group are carcinomas.

Musculoskeletal Tissues Neogenesis Soft tissues comprise a diverse family of cells and tissues whose major function is to provide form and support to the body and organs as well as to connect and anchor parts. Bone and cartilage are specialized tissues that have biphasic growth patterns. Rapid endochondral growth of the entire skeletal system occurs in the first 6 years of life, followed by quiescence between ages 6 and 12 years, and then a second spurt in adolescence between ages 12 and 18 years. It is essential to appreciate the amplification and polarity of bone growth because endochrondral bone growth varies throughout the skeleton. Osteoblasts and osteoclasts remain active in adults and remodel bone throughout life. Muscle tissue is classified as skeletal, cardiac, and smooth muscle. Striated muscle constitutes the bulk of body mass. In the 4-week embryo, mesenchymal cells at genetically predetermined sites proliferate, elongate, and differentiate into myoblasts. At 4 weeks myotubules synthesize 2 sets of myofilaments, and by 20 weeks myofilaments continue to proliferate and arrange in alternating, overlapping bonds so muscle fibers have a cross-striated appearance. The muscle growth spurts follow bone growth spurts and are biphasic. It is important to note that after muscle tissue is mature, it does not have the capability to proliferate more cells, is in a relatively fixed postmitotic state, and enlarges by hypertrophy, not hyperplasia. Neoplasia Osteosarcoma is the most common primary malignant bone tumor in children and derives from bone-forming mesenchyme. The peak incidence is in the second decade (adolescent growth spurt) and follows the amplification and polarity of skeletal growth (Fig. 1B). The bones that grow the most will be most vulnerable to become neoplastic (Fig. 5). Thus, osteosarcoma usually occurs in the metaphyses of long bones, especially around the knee joint. The highest peak of osteosarcoma occurs during pubertal years with girls having an earlier peak than boys, coincident with the rapid growth of long bones. Inactivation of the retinoblastoma gene, which is a tumor suppressor gene, may be important in the development of osteosarcoma. A common karyotypic change is deletion of the short arm of chromosome 17, where another tumor suppressor gene known as p53 is located. Osteosarcomas often arise in long-term survivors of heritable retinoblastoma and develop within or outside of radiation fields. Ewing sarcoma is the second most common childhood primary bone tumor. It tends to be bimodal in onset, that is, 30% occur in children aged ⬍ 10 years and 40% in the pubertal age range. There is debate as to the cell of origin, that is, endothelial cells, neural crest progenitors with PNET features. Again, there is a characteristic genetic defect: the 11:22 chromosomal translocation. This genetic rearrangement is detectable in 86%-90% of tumors. Primary sites tend to be

Figure 5 Skeletal distribution of primary osteocarcomas in patients treated on the Neoadjuvant Cooperative Osteo-sarcoma Study Group protocols of the Cooperative German-Austrian-Swiss Osteosarcoma Study Group. (Modified from J Clin Oncol 20:776-790, 2002.)

the pelvis, femur, rib, tibia, and humerus, where bone remodeling is very active. Rhabdomyosarcoma (RMS) is a highly malignant neoplasm arising from embryonic mesenchyme differentiating into striated muscle. RMS can arise anywhere and accounts for 3.5% of pediatric tumors and favors children aged ⬍ 15 years, with 70% occurring before 10 years and a peak at 1-4 years (35%), and another in adolescence (33%). As seen in Figure 6, the frequency of RMS largely coincides with the growth phases of muscle development. Favored sites are extremities (16%), but can arise anywhere, for example, genitourinary, parameningeal, H&N, and orbital sites. Cytogenetics include studies in myogenesis, tumor suppressor genes, translocations, that is, FKHR gene on chromosome 13 or Pax-3 gene. These translocations are present in 70% of children. Embryonal RMS is often because of heterozygosity at 11p1.5, suggesting the presence of a tumor suppressor gene at this site. Other soft tissue sarcomas comprising 3.5% of tumors include a large variety of cell types such as PNETs, neurofibrosarcomas, and synovial sarcomas. Most of these pediatric sarcomas have associated gene defects, that is, tumor suppressor genes 17q11, p53. Both RMS and non-RMS soft tissue sarcomas occur as part of the familial Li–Fraumeni syndrome.

Gonadal Germ and Stromal Cell Tumors Neogenesis Pediatric GCTs are both gonadal and extragonadal, and are related to abnormal migration of primordial germ cells

P. Rubin et al

10

Figure 6 Clinical features of rhabdomyosarcoma from IRS-I, IRS–II, and IRS-III pooled data. Age at presentation. (Modified from Principles and Practice of Pediatric Oncology. Philadelphia, PA, Lippincott, 1997, pp 803.)

(PGC); therefore the tumors can be ectopically located. After the PGCs assume positions in the germinal ridges, they migrate along these structures and develop into primitive testis and ovaries. GCTs are believed to arise from totipotent PGCs that escape normal control. Neoplasia Malignant GCTs account for 7% of all childhood neoplasms and occur in the ovaries, testes, sacrococcygeal region, vagina, retroperitoneum, pelvis, omentum, and even mediastinal areas. The main types are dysgerminoma, embryonal carcinoma, yolk-sac tumor, choriocarcinoma, and teratoma. There is a bimodal distribution in children aged ⬍ 3 years and a second peak in children aged ⬎ 12 years in the pubertal growth spurt (Fig. 1B). Among primary ovarian tumors in childhood, 70% are GCTs. Testicular GCTs in puberty tend to be yolk-sac tumors. Extranodal GCTs tend to be in the midline, consistent with migration patterns. In conclusion, the dual nature of neogenesis and neoplasia has been unequivocably demonstrated by the correlation of the surge in cancer incidence, with the growth spurt of the corresponding organ system. The peak years of an organ system’s increase in size correlate with peak years of cancer incidence in pediatric oncology. The association of genetic aberrations and mutations results in both birth defects and neoplastic disease. Cancers of epithelial origin are the least common among childhood malignancies as compared with adults.

The Epithelial Tissues and the Common Adult Malignancies Neogenesis Epithelial tissues tend to line hollow lumen and glands. The epithelial lining increases as organs increase in size, but is the smallest component of the gastrointestinal, respiratory, and genitourinary systems.

Neoplasia Unlike the child, the epithelial lining is the most mitotically active cell compartment in adults and therefore subject to malignant transformation by toxins, viruses, and even radiation. Thus, the epithelium rarely transforms into cancer during childhood because a multistep process, accompanied by an accumulation of numerous molecular changes, is generally required for a cell to transit to malignancy. However, epithelial cancers comprise a large proportion of malignancies in older individuals. As previously discussed, aging tissues experience progressive degenerative changes and gradually lose proliferative homeostasis because of several mechanisms, including apoptosis and senescence.1 Taken together, these processes explain the steeply increasing incidence rate of epithelial cancers with age (Fig. 1A). Factors that are widely recognized as contributing to adult epithelial cancers include tobacco use, alcohol intake, reproductive and other lifestyle factors (diet, obesity, physical inactivity), and oncogenic viruses. A partial list follows: ● ● ● ● ●



● ●

Tobacco Dietary factors such as excess alcohol and fat, and fruit and vegetable deficiencies Chemical carcinogens such as benzene Physical factors such as ultraviolet light, asbestos, and radiation Viruses, including both ribonucleic acid (eg, human Tlymphotrophic virus [HTLV]-1 and HTLV-2, HIV, hepatitis C) and DNA (hepadnaviruses papillomaviruses, Epstein–Barr virus, Kaposi’s sarcoma–associated herpesvirus, polyomaviruses) Inflammation, compromising tumor immune surveillance, promoting tumorigenesis, and altering the microenvironment Obesity and physical inactivity Cancer susceptibility syndromes

Tobacco smoking and alcohol intake are the major causes of epithelial tumors.18,19 About 30% of all cancer deaths are due to smoking,20 a factor attributed to the high mortality rates for cancers of the lung, larynx, esophagus, lip, oral cavity, and pharynx. Tobacco use is also associated with significantly increased risk of cancer in the stomach, pancreas, bladder, uterine cervix, and kidney.21 Alcohol also plays a significant role in cancers of the oral cavity and pharynx, larynx, esophagus, liver, large bowel, and female breast.22 Diet, obesity, physical inactivity, reproductive, and other lifestyle risk factors have a sizable effect on site-specific cancer risk in adults, including cancers of the upper and lower digestive tract, female breast, and reproductive organs.23-26 Combinations of influences may work together to increase site-specific cancer risks. For example, tobacco, alcohol, and low consumption of fruits and vegetables heighten the risk of upper aerodigestive tract tumors. Similarly, physical inactivity, caloric excess, and obesity contribute to cancer of the large bowel, as well as to hormone-dependent tumors (breast, uterine corpus, ovary, and prostate). The major human viral-associated carcinomas include those of uterine cervix (human papillomavirus), liver (hepatitis B virus and hepatitis C virus), and nasopharynx (Ep-

Cancer genesis across the age spectrum stein–Barr virus).27 It is estimated, however, that only about 7% of incident cancers in the United States are due to oncogenic infections.28 As depicted in Figure 1A, the rising incidence of most adult cancers with increasing age is likely to be associated with the previously discussed toxic exposures and failure to repair injury. Reasons for the plateau or decrease in tumor incidence in the very elderly are discussed in the preceding paragraphs. Epithelial cancers rarely occur during the first 2 decades of life, and include adrenal cortical carcinoma, H&N carcinoma, Epstein–Barr virus–associated nasopharyngeal carcinoma, and esophageal carcinoma (which may develop after chemical burns, for example, lye ingestion). Lung cancers that develop in children tend to consist of pleuropulmonary blastomas, mesenchymal sarcomas, and mesotheliomas, whereas stomach tumors comprise lymphomas, leiomyomas, leiomyosarcomas, and leiomyoblastomas. Similarly, colon, intestinal, and pancreatic cancers are rare in children as compared with lymphomas, carcinoids, PNETs, and pancreatic blastomas. Breast carcinoma is less common than cystosarcoma phylloides.

11

4. 5. 6. 7.

8. 9. 10. 11. 12.

13. 14.

Conclusions Cancer genesis across the age spectrum is a complex, multifactorial process. Cancer is not a single disease, and different cancer cells (and stem cells) may demonstrate aberrant function in various ways. Mutations in DNA, some random and some explained by exogenous insults, accompanied by changes in the tissue microenvironment, generally precede the onset of aberrant replication and apoptosis. Recently, Naxerova et al29 of the Harvard-MIT Division of Health Sciences and Technology categorized cancer into 3 groups that correlate with different embryonic stages on the basis of evidence that genetic programs normally active only during human development are reactivated during tumorigenesis. Regardless, in childhood we have demonstrated the dual nature of neogenesis and neoplasia: the peak years of an organ system’s increase in size correlates with peak years of cancer incidence. Conversely, in most adult-onset cancers, it is the exposure to exogenous toxins, the failure of maintenance and repair, and eventually cell loss with senescence that explains many malignancies. Additional basic science investigations and epidemiologic studies will assist in our understanding of the mechanisms that underlie the notable difference in sitespecific cancer incidence according to age.

Acknowledgment The authors thank David Check of the Biostatistics Branch, Division of Cancer Epidemiology and Genetics, National Cancer Institute, for figure development.

References 1. Campisi J, Sedivy J: How does proliferative homeostasis change with age? What causes it and how does it contribute to aging? J Gerontol A Biol Sci Med Sci 64:164-166, 2009 2. Garcion E, Naveilhan P, Berger F: Cancer stem cells: Beyond Koch’s postulates. Cancer Lett 278:3-8, 2009 3. Park CY, Tseng D, Weissman IL: Cancer stem cell-directed therapies:

15.

16.

17.

18. 19. 20. 21.

22.

23. 24.

25. 26. 27.

28.

29.

Recent data from the laboratory and clinic. Mol Ther J Am Soc Gene Ther 17:219-230, 2009 Brooks AL: Paradigm shifts in radiation biology: Their impact on intervention for radiation-induced disease. Radiat Res 164:454-461, 2005 Rugh R: The impact of ionizing radiation on the embryo and fetus. Am J Roentgenol 89:182-190, 1963 DiPaolo JA, Kotin P: Teratogenesis-oncogenesis: A study of possible relationships. Arch Pathol 81:3-23, 1966 Durante F: Nesso fisio-patologico tra la struttura dei nei materni e la genesi di alcuni tumori maligni. Arch Memor Observ Chir Prat 11:217, 1874 Cohnheim J: Lecture on General Pathology, vol. 2. London, UK, The New Syndenham Society, 1889 van Den Hoff MJ, van Den Eijnde SM, Viragh S: Programmed cell death in the developing heart. Cardiovasc Res 45:603-620, 2000 Chan WY, Lorke DE, Tiu SC: Proliferation and apoptosis in the developing human neocortex. Anat Rec 267:261-276, 2002 Hanahan D, Weinberg RA: The hallmarks of cancer. Cell 100:57-70, 2000 Postovit LM, Magarayan NV, Seftor EA: Human embryonic stem cell microenvironment suppresses the tumorigenic phenotype of aggressive cancer cells. Proc Natl Acad Sci U S A 105:4329-4334, 2008 Malanchi I, Huelsken J: Cancer stem cells: Never Wnt away from the niche. Curr Opin Oncol 21:41-46, 2009 Wang Z, Li Y, Banerjee S, et al: Emerging role of Notch in stem cells and cancer. Cancer Lett 279:8-12, 2009 Ries LAG, Melbert D, Krapcho M (eds): SEER Cancer Statistics Review, 1975-2005. Bethesda, MD, National Cancer Institute. Available at: http://seer.cancer.gov/csr/1975_2005/ (based on November 2007 SEER data submission, posted to the SEER web site, 2008). Accessed April 2009 SEER Statistics Database: Incidence—SEER 9 Registries, Limited-Use, November 2007 Sub (1973-2005) (Katrina/Rita Population Adjustment)⫺ Linked To County Attributes—Total USA, 1969-2005 Counties, National Cancer Institute, DCCPS, Surveillance Research Program, Cancer Statistics Branch, released April 2008. Accessed November 2007 Surveillance Research Program, National Cancer Institute SEER Statistics software. Available at: http://www.seer.cancer.gov/seerstat. Version 6.5.1. Vineis P, Alavanja M, Buffler P: Tobacco and cancer: Recent epidemiological evidence. J Natl Cancer Inst 96:99-106, 2004 Boffetta P, Hashibe M, LaVecchia C: The burden of cancer attributable to alcohol drinking. Int J Cancer 119:884-887, 2006 Doll R, Peto R: The causes of cancer: quantitative estimates of avoidable risks of cancer in the US today. J Natl Cancer Inst 66:1191-1308, 1981 Thun MJ, Henley SJ: Tobacco, in Schottenfeld D, Fraumeni JF Jr (eds): Cancer Epidemiology and Prevention. New York, NY, Oxford University Press, 2006, pp 217-242 Marshall JR, Freudenheim J: Alcohol, in Schottenfeld D, Fraumeni JF Jr (eds): Cancer Epidemiology and Prevention. New York, NY, Oxford University Press, 2006, pp 243-258 Henderson BE, Feigelson HS: Hormonal carcinogenesis. Carcinogenesis 21:427-433, 2000 Calle EE, Rodriguez C, Walker-Thurmond K: Overweight, obesity, and mortality from cancer in a prospectively studied cohort of US adults. N Engl J Med 348:1625-1638, 2003 Key TJ, Schatzkin A, Willett WC: Diet, nutrition and the prevention of cancer. Pub Health Nutr 7:187-200, 2004 Pike MC, Pearce CL, Wu AH: Prevention of cancers of the breast, endometrium and ovary. Oncogene 23:6379-6391, 2004 Mueller NE, Birmann BM, Parsonnet J: Infectious agents, in Schottenfeld D, Fraumeni JF Jr (eds): Cancer Epidemiology and Prevention. New York, NY, Oxford University Press, 2006, pp 507-548 Parkin DM, Pisani P, Munoz N: The global health burden of infection associated cancers, in Newton R, Beral V, Weiss RA (eds): Cancer Surveys: Infections and Human Cancer. Plainview, NY, Spring Harbor Laboratory, 1999, pp 5-33 Naxerova K, Bult CJ, Peaston A: Analysis of gene expression in a developmental context emphasizes distinct biological leitmotifs in human cancers. Genome Biol 9:R108, 2008