Page 314 - 20dynamics of cancer

CELL LINEAGE HISTORY 299

Tavare 2006). For example, Kim et al. (2006) measured mitotic age in
hair follicles by the frequency of CpG methylation, as I discussed in an
earlier section.
Two studies report the age-specific frequency of methylation in the
colon and endometrium (Yatabe et al. 2001; Kim et al. 2005). The colon
shows a continuous rise in methylation frequency with age (Figure 14.5a).
That steady rise in methylation supports the usual view of the colon as
a continuously renewing tissue throughout life.
By contrast, methylation of the endometrium increases sharply to the
age of menopause, then levels off through the remainder of life (Fig-
ure 14.5b). The early-life rise in endometrial methylation corresponds
to the period of menstrual cycling and frequent tissue renewal. The late-
life plateau corresponds to the period of reproductive quiescence and
limited turnover of reproductive tissues.
Two further observations support the hypothesis that methylation
correlates with mitotic age. Obese women typically have higher estro-
gen levels and greater reproductive tissue renewal than do lean women;
obese women had correspondingly higher methylation levels of endome-
trial tissue than lean women (Figure 14.5b). Women with two or fewer
children typically have more lifetime menstrual cycles than do women
with three or more children; those women with fewer children and more
menstrual cycles had correspondingly higher levels of methylation than
those women with more children (Figure 14.5b).

MITOTIC AGE AND INCIDENCE IN FEMALE REPRODUCTIVE TISSUES
If mitoses do in fact drive progression, then the patterns of age-
specific mitosis should correspond to patterns of age-specific incidence.
Pike et al. (1983) argued that reduced mitotic rate of the breast after
menopause causes the observed drop in the slope of the age-specific
incidence of breast cancer later in life. Pike et al. (2004) updated the
analysis to include the slowing rate of increase in cancer of the breast,
ovary, and endometrium later in life. In Pike’s formulation of the theory,
incidence increases with mitotic age, so the rise in incidence for female
reproductive tissues slows later in life as mitosis slows.
I use log-log acceleration (LLA) to measure the change in incidence
with age. Figure A.8 shows plots of LLA for ovarian cancer. Those LLA
curves follow the declining acceleration through the later part of life
described by Pike. Figure A.2 shows that breast cancer also has declining