BA21 - Understanding the role of sex hormones in colorectal cancer

Investigator Names and Contact Information

Marc Gunter (m.gunter@imperial.ac.uk)

Introduction/Intent

This application proposes a prospective nested case-control study of endogenous sex hormones and their associations with risk of incident colorectal cancer. Specifically, we will measure circulating levels of estradiol, estrone, progesterone and sex hormone binding globulin (SHBG) in 400 incident colorectal cancer cases and 800 controls (matched by time in the WHI-CT, age, race, and study arm), randomly selected from the non-treatment arms of the WHI-CT; i.e., those randomized to placebo in the E and E+P trials, or who opted to participate only in the dietary and calcium intervention trials. Each group of women met the eligibility criteria for the WHI-CT and were recruited in the same manner as those who received E+P. Multivariate conditional logistic regression will be used to assess the independent associations between each sex hormone and colorectal cancer, controlling for insulin, free IGF-I, CRP and other colorectal cancer risk factors. We note that this study was specifically designed to have adequate statistical power to assess the independent associations of each of the several somewhat correlated factors that will be measured.

The Women’s Health Initiative clinical trial (WHI-CT) found that hormonal therapy with combined estrogen and progestin (E+P), but not estrogen alone (E), significantly reduced the risk of colorectal cancer in postmenopausal women. However, the biologic mechanisms underlying the protective effect of E+P, and explaining its absence in E, are unknown. Indeed, little at all is known regarding the role of (e.g., endogenous) sex hormones in colorectal tumorigenesis.

To begin to understand the role of sex hormones in colorectal tumorigenesis, we recently examined the relation of endogenous estradiol levels with risk of incident colorectal cancer in the WHI-observational study. The study showed that high endogenous serum estradiol was significantly associated with increased (not decreased) risk of colorectal cancer in post-menopausal women who were not using hormone therapy (non-HT users). Several laboratory studies also reported that estradiol has tumorigenic effects on colorectal cells, whereas estrone (the most abundant estrogen in E and E+P) did not. Thus, estradiol and estrone could have different associations with risk of colorectal cancer. Additional studies are now needed to validate and expand our knowledge of the role of estrogens in colorectal tumorigenesis; measuring not only total estradiol, but also estrone and sex hormone binding globulin (SHBG). Furthermore, no studies to our knowledge have assessed the relation of colorectal cancer with progesterone, the factor that most obviously distinguishes E+P from E.

Determining the role of sex hormones in the development of colorectal cancer is essential to understanding the protective effects of E+P and, from a scientific perspective, will provide insight into previously uncharacterized carcinogenic pathways. Clinically, these hormonal pathways might also be exploited for novel chemopreventive and therapeutic methods. However, the synthetic progestin used in WHI, medroxyprogesterone acetate (MPA), does not reach a steady state and cannot be meaningfully measured without multiple blood specimens collected throughout the day (specimens that are unavailable in WHI, or other relevant studies).

Therefore, among women in the untreated arms of the WHI-CT, we propose a nested case-control study of incident colorectal cancer (n=400 cases, n=800 controls) and its associations with levels of estradiol, estrone, progesterone, and SHBG in serum or plasma (whichever is most available). Testing for insulin, free/unbound insulin-like growth factor (IGF)-I, and C-reactive protein (CRP) will also be conducted, to control for these possible colorectal cancer risk factors in our models. This proposal satisfies two main goals of the BAA, namely, validating a novel observation made in the WHI (i.e., the relation of estradiol with colorectal cancer) and, secondly, elucidating the biologic pathways that may underlie the protective effect of E+P.

Major Hypotheses: The incidence of colorectal cancer will be positively associated with levels of estradiol (particularly free estradiol), while estrone levels will have no association with colorectal cancer risk. Progesterone levels, in contrast, will be inversely associated with colorectal cancer. The above results will be not be meaningfully affected by statistical adjustment for levels of insulin, free IGF-I, CRP, or for other colorectal cancer risk factors.

Specific Aim:

To validate endogenous estradiol and (for the first time) evaluate endogenous progesterone as risk markers for colorectal cancer in non-HT users (the non-treatment arms of the WHI-CT), by:

a. Determining the independent associations of estrogen levels (total estradiol, free estradiol, estrone; each assessed in separate models) with risk of incident colorectal cancer - controlling for serologic (insulin, free/unbound IGF-I, CRP), and other risk factors for colorectal cancer.

b. Determining the independent association of progesterone levels with risk of colorectal cancer, controlling for estrogen (whichever parameter is most strongly related to colorectal cancer - above), additional serologic, and other risk factors for colorectal cancer.

Results/Findings

For a complete, up-to-date list of WHI papers related to this ancillary study, please use the searchable Papers section of this website.