Page 170 - 20dynamics of cancer

GENETICS OF PROGRESSION 155

;MPH X]TI
,IXIVS^]KSXI

*VEGXMSR WYVZMZMRK

(E]W

Figure 8.6 Survival of wild-type TRAMP mice versus Pten heterozygous TRAMP
mice that have one Pten allele knocked out. Kwabi-Addo et al. (2001) ascribed
death in all 63 mice shown in these plots to either a large primary tumor or
to metastatic disease. Survival plots of this sort are often called Kaplan-Meier
plots.

Lab studies rarely analyze the quantitative patterns of cancer onset
in the way that I did in the previous sections. Instead, the analysis typ-
ically emphasizes the qualitative pattern of whether certain combina-
tions of mutations cause earlier or later cancer onset than do other
combinations. For example, Figure 8.6 compares the survival of two
mouse strains (Kwabi-Addo et al. 2001). One strain has the TRAMP geno-
type that predisposes mice to develop prostate cancer. The other strain
carries the same genes that predispose to prostate cancer, but also is
heterozygous at the Pten locus, with one allele knocked out. Pten mu-
tations are common in many cancers, including cancers of the prostate.
The figure shows that the Pten heterozygotes progress more rapidly to
cancer.
Experimenters usually plot results from these studies as the fraction
of mice surviving to a particular age, as in Figure 8.6. In this section,
I show how to transform such data into age-specific rates of cancer in-
cidence, allowing comparison of relative rates for different treatments.
This transformation to age-specific rates allows one to test particular
hypotheses about the dynamics of cancer onset with the limited sample
sizes typical of lab studies. I illustrate the method by analyzing the age
of cancer onset in different DNA mismatch repair genotypes (Frank et al.
2005).