Professional Documents
Culture Documents
There has been and will continue to be a dramatic shift in the composition of our
population with regard to age. Over the next several decades, the percentage of
the population that is older than 65 years will nearly double. This has obvious and
striking implications, particularly for health care. In anticipation of this change,
there has been increased research activity in an attempt to understand the basic
biology of aging and the mechanisms whereby older individuals become susceptible to disease. The authors reviewed MEDLINE data from the past 2 decades
and key articles from the literature to develop a comprehensive overview of the
background and current status of research in biomedical gerontology. They found
that research in the mechanisms of aging has expanded strikingly during the past
20 years, and that there is now a more precise understanding of age-associated
changes in cellular and molecular functions. Scientific overlap of these age-associated changes with the biology of cancer is now recognized. An appraisal of our
current understanding of aging (or lack thereof) is made in this article. Selected
key research questions are also presented, some of which are particularly germane
to cancer biology and clinical oncology. Cancer 1997;80:128493.
q 1997 American Cancer Society.
KEYWORDS: aging, theories of aging, cancer and aging, immune senescence, free
radicals, telomeres, immunosurveillance.
Presented at Oncology Geriatric Education Retreat: Integrating Geriatrics into Oncology Training, San Juan, Puerto Rico, February 2126,
1997.
Address for reprints: William B. Ershler, M.D.,
Gerontology Research Center, National Institute
on Aging, National Institutes of Health, 4940
Eastern Avenue, Baltimore, MD 21224.
Received May 8, 1997; revision received July
7, 1997; accepted July 7, 1997.
ellular senescence and malignant transformation share certain basic pathways. In fact, in examining genes that control cellular
proliferation and cell death, molecular biologists have discovered that
mutations or alterations in the functions of some of these genes can
result in immortalized cells with malignant properties (i.e., growth
factor independent proliferation and invasiveness). Thus, at the cellular level, aging and cancer can be considered distinct steps on a
spectrum of cell behavior. Recently, there has been an increased
awareness of this paradigm, and those who study the basic sciences
of aging and cancer are opening a dialogue that has been mutually
beneficial.
In this review, currently prevalent theories of aging are discussed
and presented in the context of their applicability to cancer biology.
These include theories regarding programmed (genetic) aging, protein
and DNA glycation, and free radicals. Experimental models are reviewed herein, with mention of the influence of dietary restriction on
both aging and cancer development. Several key research questions
are presented, to which both oncology and gerontology perspectives
are critical. These include questions about the increase in cancer
incidence with advancing age and changes in tumor aggressiveness
that occur in older hosts. Other questions, about the frequency of
multiple primary cancers in older individuals and the apparent resis-
/ 7b71$$ge4D
09-10-97 11:25:47
cana
W: Cancer
/ 7b71$$ge4D
09-10-97 11:25:47
cana
1285
W: Cancer
1286
caloric intake is restricted are assuming a more youthful phenotype in a variety of physiologic measures,25,26
it is clearly too early to predict whether maximum
survival will be affected.
/ 7b71$$ge4D
09-10-97 11:25:47
cana
and disease and, by necessity, have selected appropriate animal models for their investigations. A general
rule, however, is that short-lived strains, particularly
those with a genetic susceptibility to life-threatening
diseases, are not good models for research in aging.
In contrast, factors that influence maximum survival
in robust hybrid strains probably do so by interfering
with primary aging processes; thus, these strains make
excellent models.
W: Cancer
Theories
Intrinsic/Stochastic
Somatic mutation40,41
Intrinsic mutagenesis45
Impaired DNA repair46
Error catastrophe47
Ionizing radiation40,41,43,49
Free radical damage52,53
Neuroendocrine91
Immune68
Extrinsic/Stochastic
Genetically determined
Theories of aging
Providing a rational, unifying explanation for the aging
process has been the subject of a great number of
theoretic expositions. Yet, no single proposal has sufficed to account for the complexities observed. Studies
that present theories of aging are listed in Table 1.
That genetic controls are involved seems obvious
when one considers that life span is highly species
specific. For example, mice generally live approximately 2 years and humans approximately 90 years.
However, the aging phenomenon is not necessarily a
direct consequence of the primary DNA sequence.
Mice and bats, for instance, have a 0.25% difference
in their primary DNA sequence, but bats live for 25
years 10 times longer than mice. Thus, regulation
of gene expression seems likely to be the source of
differences in longevity between species.
Although within a species there is considerable
variation in longevity, this variability is much less
among inbred strains or monozygotic twins than
among dizygotic twins or nontwin siblings. Furthermore, various genetically determined syndromes have
remarkable (albeit incomplete) features of accelerated
aging. These include Hutchinson Gilford syndrome
(early onset progeria), Werners syndrome (adult onset
progeria), and Downs syndrome.37 Although no progeria syndrome manifests a complete phenotype of
advanced age, the identification of the genes responsible for these particular syndromes is beginning to pay
dividends by providing clues to the molecular mechanisms involved in the aging process. For example, Wer-
/ 7b71$$ge4D
09-10-97 11:25:47
cana
1287
ners syndrome is now known to be caused by mutations in a single gene on chromosome 8 that encodes
a protein containing a helicoid domain.38,39 The future
functional characterization of this specific protein will
undoubtedly increase our level of understanding of
the aging process.
Examination of aging in yeast has also been informative with regard to the genetic controls of aging.
These single-cell organisms follow the replicative limits of mammalian cells, and it has been observed that
life span is related to silencing large chromosomal
regions. Mutations in these silencing genes lead to
increased longevity.40
Thus, if there are certain genes that regulate normal aging or at least are associated with the development of a phenotype in the aged, it stands to reason
that acquired damage to those genes might influence
the rate of aging. Over the years, several theories have
been proposed that relate to this supposition. In general, they hypothesize a stochastic accumulation of
damage, to either DNA or protein, that leads eventually to dysfunctional cells, cell death, subsequent organ dysfunction, and ultimately organism death.
Prominent among these is the somatic mutation theory,41 which predicts that genetic damage (from background radiation, for example) accumulates, produces
mutations, and results in functional decline. A variety
of refinements to this theory have been suggested, invoking the importance of mutational interactions,42
transposable elements,43 and changes in DNA methylation status.44
A related hypothesis is Burnets intrinsic mutagenesis theory,45 which proposes that spontaneous or endogenous mutations occur at different rates in different species and that this accounts for the variability
observed in life span. Closely related to this notion
is the DNA repair theory.46 Initially, there was great
excitement about this, as it was found that long-lived
animals had demonstrably greater DNA repair mechanisms than shorter-lived species.46 However, longitudinal studies of particular species have not revealed
that repair mechanisms decline consistently with age.
This, of course, does not rule out the possibility that
repair of certain specific and critical DNA lesions is
altered with advancing age. We now understand that
there are multiple DNA repair mechanisms, including
base excision repair, transcription-coupled repair,
and, as was discovered most recently, even DNA repair
mechanisms based in mitochondria. Disorders involving one or a subset of repair mechanisms could lead
to accumulation of DNA damage and dysfunction.
In yet another intrinsic/stochastic model, the error catastrophe theory proposed by Orgel,47 it was suggested that random errors in protein synthesis occur
W: Cancer
1288
and that when the proteins involved are those responsible for DNA or RNA synthesis, there is resultant DNA
damage with consequences for daughter cells. Although this model has appeal, there has been no reported evidence that protein synthesis machinery becomes impaired or inaccurate with advancing age.
However, a candidate protein that may eventually be
shown to be so affected is telomerase. This critical
enzyme is comprised of protein and an RNA template;
the template is necessary for maintaining telomere
length and cell replicative potential. As cells become
senescent in vitro, telomerase activity declines, telomeres shorten, and ultimately replicative potential
is lost.48
Evidence that exogenous factors are involved in
the acquisition of age-associated damage to DNA and
protein is derived from a number of observations,
many of which are circumstantial or correlative but
nonetheless provocative. It now appears that the accumulation of abnormal protein within senescent cells,
as predicted by the error catastrophe theory, actually
reflects postranslational events, such as oxidation or
glycation and resultant cross-linking. There is theoretic appeal in the concept that key proteins, such
as collagen or other extracellular matrix proteins, and
DNA become dysfunctional with age due to the impairment produced by these cross-links.49 51
One mechanism that produces cross-links is
called glycation, the nonenzymatic reaction of glucose
with the amino groups of proteins. Presumably, glycation would occur more readily in the presence of
higher serum levels of glucose; thus, this theory fits
well with the observed age-associated dysregulation
of glucose metabolism and prevalent hyperglycemia
in geriatric populations. Of course, this also points out
its deficiency as a unifying mechanism, as there is no
question that individuals with well-maintained glucose levels throughout their life span will still be subject to acquired changes that are typical of aging.
Another mechanism held responsible for crosslinking is the damage produced by free radicals, and
this forms the basis of the free radical hypothesis initially promoted by Harman.52,53 This theory suggests
that aging is the result of DNA and protein damage
(e.g., mutagenesis or cross-linking) by atoms or molecules that contain unpaired electrons (free radicals).
These highly reactive species are produced as by-products of a variety of metabolic processes and are normally inhibited by intrinsic cellular antioxidant defense mechanisms. If free radical generation increases
with age, or the defense mechanisms that scavenge
free radicals (e.g., glutathione) or repair free radical
damage decline, the accumulated free radical damage
may account for altered DNA and protein function.
/ 7b71$$ge4D
09-10-97 11:25:47
cana
Evidence to support this widely held notion is incomplete. It is known that free radical generation in mammals correlates inversely with longevity; 54 similarly,
levels of free radical inhibiting enzymes, such as superoxide dismutase, have been shown to be higher in
species with longer life spans.55 However, efforts at
enhancing antioxidant mechanisms with dietary vitamin E have resulted in only a modest enhancement
of median survival in mice and no affect on maximum
life span.56
Recently, there has been much attention focused
on mitochondrial function in the context of free radical damage, because oxidative metabolism and the
production of reactive oxygen species mostly occur in
these organelles. Although mitochondrial DNA codes
for antioxidant enzymes in addition to enzymes involved in energy production, it is currently believed
that energy production declines with age due to mitochondrial DNA damage by those reactive products.
Indeed, mitochondrial damage has been shown to increase with age in experimental models,57 59 and the
shortened survival of knockout mice that are deficient
in mitochondrial antioxidant enzymes has supported
the potential importance of this mechanism.60
The most compelling data to date in support of
the free radical hypothesis come from the experiments
of Orr and Sohal, in which transgenic Drosophila producing enhanced levels of superoxide dismutase and
catalase had a maximum survival 33% greater than
controls.23 Furthermore, it is known that flies produce
high levels of free radicals associated with their impressive metabolic requirements and that their survival is enhanced dramatically when their ability to
fly is experimentally hindered.61 However, the idea of
generalizing these findings has been questioned. Some
have criticized the transgenic Drosophila experiment,
claiming that the controls were rather short-lived. Furthermore, transgenic mice overexpressing free radical
scavenging enzymes have produced very modest effects on life span.62 Thus, the conclusion that augmentation of free radical scavenging mechanisms increases
longevity cannot be considered a proven fact.
From a different perspective, there is also very
good evidence implicating a nonrandom, perhaps genetically regulated endogenous mechanism in the
aging process. For example, the neuroendocrine theory suggests that the decrements in neuronal and associated hormonal function are central to aging. It has
been suggested that age-associated decline of hypothalamic-pituitary-adrenal axis function results in a
physiologic cascade leading ultimately to the frail
phenotype. This hypothesis is appealing because it is
well established that this neuroendocrine axis regulates much of the development and also the involution
W: Cancer
of ovarian and testicular function. Furthermore, ageassociated declines in growth hormone and related
factors,63 dehydroepiandrosterone,64 and secondary
sex steroids65 have been implicated in age-associated
impairments, including a reduction in lean body mass
and bone density. Furthermore, pharmacologic reconstitution using these or related hormones has met with
some success in reversing age-associated functional
decline.66,67
Similarly, it has been argued that involution of
the thymus gland and subsequent decline in immune
function is a key regulator of aging.68 The argument is
based on the observation that the decline in immune
function occurs in all mammalian species but occurs
later in those with longer survival.69 Furthermore, dietary restriction is associated with maintained thymic
mass and measurable immune function as well as prolonged survival, suggesting an association of a decline
in immunity with primary aging processes.
This possibility is highlighted by the observation
that differences in the maximum survival of different
mouse strains has been associated with specific alleles
in the major histocompatibility complex, which in turn
code for immunologic determinants.70 This hypothesis, although not without its appeal, is not widely accepted as a major explanation for aging. Perhaps this
relates to the fact that biologic aging is a universal
phenomenon and certain features of it are held in
common, even among organisms with primitive or no
immune function. (The same could also be said for the
neuroendocrine theory.) It is obvious that the immune
system is of great importance in minimizing the
chance of early death, particularly from infectious diseases. Accordingly, median life span is clearly influenced by competent immunity. However, immunologic reconstitution of middle-aged or old animals has
not been shown to prolong survival.71
/ 7b71$$ge4D
09-10-97 11:25:47
cana
1289
TABLE 2
A Sampling of Candidate Biomarkers of Aging
Subjects
Markers
Cells in Culture
b-galactosidase36
telomere length48
marrow stromal colony size92
protein carbonyl content93
mitochondrial H2O2 release61
striatal dopamine reeceptors94
Ca// ionophore lymphocyte response95
tail tendon fiber strength96
mitochondrial DNA damage57
presbyopia97
fingernail growth rate98
interleukin-6 level99
glycation end products103
mitochondrial DNA damage59
b-galactosidase in skin biopsy36
lenticular glutathione100
presbyopia101
DNA unwinding rate102
Flies
Mice, Rats
Monkeys
Humans
W: Cancer
1290
paraproteinemia of aging. Typically, this is accomplished by examination of bone marrow, skeletal Xrays, and renal function.84
Other indications of dysregulated immune function are the alterations in certain key cytokines that
are measured in plasma, culture supernatants, or the
appropriate tissue microenvironment. Notably and
consistently, interleukin (IL)-2 levels and function decrease with age,85 whereas IL-6 levels increase.86 The
decline in IL-2 may account for a significant component of the measured decline in T-cell function, and
the increased IL-6 has been implicated in the pathogenesis of certain age-associated diseases, including
osteoporosis, Alzheimers disease, and cancer.87
/ 7b71$$ge4D
09-10-97 11:25:47
cana
This figure illustrates the proposed counterimmune forces that might influence tumor growth and explain the apparent reduction in tumor growth
that occurs with age. Shaded figures of different sizes represent tumors
growing at different rates. Tumors 2 and 4 represent weakly antigenic
tumors, whereas 1 and 3 represent strongly antigenic tumors. The two
figures at the top represent tumors in an immunocompetent (or young)
host, wheras the figures below represent tumors in an immunodeficient
(or older) host. The strength of immune-facilitating forces (for each tumor,
the set of arrows within the tumor) and of immunosuppressive forces (for
each tumor, the external arrow on the left side) are represented by the
sizes of the arrows. The strongly antigenic tumors induce strong immunosuppressive responses only in immunocompetent hosts, and tumor growth
is therefore inhibited. Immunodeficient hosts produce a weak immunosuppressive response to a strongly antigenic tumor, and tumor growth is
correspondingly greater (Tumor 2 vs. Tumor 1). Weakly antigenic tumors
induce little immunosuppressive response in either immunocompetent or
immunosuppressed hosts. The immunocompetent host will produce a
greater immune-facilitating response (see text), and tumor growth will be
greater than in the immunosuppressed host (Tumor 2 vs. Tumor 4). This
hypothesis predicts that in an immunodeficient (aged) host with a weakly
antigenic tumor, the tumor growth would be slow compared with the
growth of the same tumor in an immunocompetent young host.
ation, and genetic control are among the many common research fronts. With the aging of the population
and the increased frequency of cancer, a research
agenda that relates both to primary aging and cancer
biology is of obvious importance. Among the important basic clinical research questions to address
are the following:
1. What determines whether the accumulation of genetic damage in a particular cell leads to death,
growth arrest/senescence, or neoplasia?
2. Which genes regulate cellular aging, and do they
relate to organismal aging?
3. To what extent does free radical damage contribute
W: Cancer
4.
5.
6.
7.
8.
REFERENCES
1.
2.
3.
4.
5.
6.
7.
8.
9.
10.
11.
12.
13.
14.
15.
/ 7b71$$ge4D
09-10-97 11:25:47
cana
1291
W: Cancer
1292
/ 7b71$$ge4D
09-10-97 11:25:47
cana
60. Li Y, Huang YY, Carlson EJ, Melov S, Ursell PC, Olson JL, et
al. Dilated cardiomyopathy and neonatal lethality in mutant
mice lacking manganese superoxide dismutase. Nature
1995;11:37681.
61. Sohal RS. Hydrogen peroxide production by mitochondria may
be a biomarker of aging. Mech Ageing Dev 1991;60:18998.
62. Epstein CJ, Avraham KB, Lovett M. Transgenic mice with
increased Cu/Zn-superoxide dismutase activity: animal
model of dosage effects in Downs syndrome. Proc Natl Acad
Sci U S A 1987;84:80448.
63. Harris TB, Roubenoff RR, Langlois J, Hannan M, Havlik R,
Wilson P. Association of insulin-like growth factor-I with
body composition, weight history, and past health behaviors
in the very old: the Framingham Heart Study. J Am Geriatr
Soc 1997;45:1339.
64. Birkenhager-Gillesse EG, Derkson J, Lagaay AM. Dehydroepiandrosterone sulfate (DHEAS) in the oldest old, aged 85
and over. Ann N Y Acad Sci 1994;719:5432.
65. Rudman D, Drinka PJ, Wilson CR, Scherman F, Cuisinier
MC, Schults S. Relations of endogenous anabolic hormones
and physical activity to bone mineral density and lean body
mass in elderly men. Clin Endocrinol 1994;40:65361.
66. Rudman D, Feller AG, Nagraj HS, Gergans GA, Lalitha PY,
Goldberg AF, et al. Effects of human growth hormone in
men over 60 years old. N Engl J Med 1990;323:16.
67. Hobbs CJ, Plymate SR, Rosen CJ, Adler RA. Testosterone
administration increases insulin-like growth facto-1 levels
in normal men. J Clin Endocrinol Metab 1993;77:7769.
68. Walford R. The immunologic theory of aging. Copenhagen:
Munksgard, 1969.
69. Makinodan T, Kay MMB. Age influence on the immune system. Adv Immunol 1980;29:287330.
70. Smith GS, Walford RL. Influence of the main histocompatibility complex on aging in mice. Nature 1977;270:7279.
71. Hirokawa K, Albright JW, Makinodan T. Restoration of impaired immune function in aging animals. I. Effect of syngeneic thymus and bone marrow grafts. Clin Immunol Immunopathol 1985;5:3716.
72. Baker GT, Sprott RL. Biomarkers of aging. Exp Gerontol
1988;23:22339.
73. Gatti RQ, Good RA. Aging, immunity and malignancy. Geriatrics 1979;25:15868.
74. Bulychev VV. Longevity, atherosclerosis and cellular immunity. Klin Med (Mosk) 1993;71:514.
75. Lehuen A, Bendelac A, Bach JF, Carnaud C. The nonobese
diabetic mouse model: independent expression of humoral
and cell mediated autoimmune features. J Immunol 1990;
144:214751.
76. Hull M, Strauss S, Berger M, Volk B, Bauer J. The participation of interleukin-6, a stress-inducible cytokine, in the
pathogenesis of Alzheimers disease. Behav Brain Res 1996;
78:3741.
77. Hull M, Fiebich BL., Lieb S, Berger SS, Volk B, Bauer J. Interleukin-6associated inflammatory processes in Alzheimers disease: new therapeutic options. Neurobiol Aging
1996;17:795800.
78. Gillis S, Kozak R, Durante M, Weksler ME. Decreased production and response to T cell growth factor by lymphocytes
from aged humans. J Clin Invest 1981;67:93742.
79. Miller RA. The aging immune system: primer and prospectus. Science 1996;273:704.
80. Stephan RP, Sanders VM, Witte PL. Stage-specific alterations
in murine lymphopoiesis with age. Int Immunol 1996;8:
50918.
W: Cancer
/ 7b71$$ge4D
09-10-97 11:25:47
cana
1293
W: Cancer