Dean Hosgood

PhD MPH

Albert Einstein College of Medicine

Assistant Professor

PLCO (Learn more about this study)

PLCO-221

Aug 23, 2016

Sex hormones, hyperinsulinemia, and the risk of hematopoietic malignancies of the B-cell lineage

Non-Hodgkin lymphoma (NHL) and multiple myeloma (MM), both arising from B-cell lymphocyte perturbations, are the first and second most common hematologic malignancies in the US, with NHL representing the 6th most common cause of cancer death. B-cell lymphomas and MM have few known risk factors beyond family history and immunosuppression. Recent data suggest that hormonal factors, including sex hormones and hyperinsulinemia, may influence the risk of NHL and MM.

Laboratory studies have shown that sex hormones can induce lymphoma in animal models4. Estrogen is involved in the differentiation, proliferation, and survival of hematopoietic cells. Furthermore, estrogen has been shown to regulate early lymphoid precursors in bone marrow and the development and maturation of B-cells in male mice, potentially via estrogen receptor signaling. Administering estrogen (either alone or with progesterone) to both male and female mice without exceeding normal physiological levels, showed that estrogen acted directly on hematopoietic stem cells, rather than indirectly by stimulating secondary signals. Menstrual and reproductive factors such as parity, age at menarche, and age at menopause have long been known to be risk factors in solid tumors, such as breast, where estrogen is involved. In humans, a pooled analysis of 9 case-control studies observed that postmenopausal women who had used hormone therapy (i.e., exogenous sex hormones) were at decreased risk of NHL. Few studies have investigated the association between menstrual and reproductive history and the risk of hematopoietic cancers. To date, the focus of previous studies has been on NHL, with previous studies suggesting that although the overall evidence is minimal, increased parity may be associated with decreased risk of this type of NHL. However, weak associations could be masked by heterogeneity between NHL subtypes, so it is important to consider these separately. For example, in the pooled analysis of InterLymph case-control studies, increasing parity was associated with decreased risk of follicular lymphoma (FL), but not diffuse large cell B-cell lymphoma (DLBCL), nor NHL overall.

Another hormone-related mechanism, insulin signaling, may also contribute to the etiology of hematopoietic malignancies. Indirect support for the concurrent involvement of sex hormones and insulin in the etiology of hematopoietic malignancies comes from the observation that obesity, which is associated with alterations in levels of these factors, has been associated with increased risk of B-cell NHL and MM in prospective cohort studies; however, direct evidence is lacking as this concurrent involvement has not been studied in a robust prospective cohort study.

We hypothesize that differences in exposure to sex hormones and insulin may affect the risk of hematopoietic malignancy later in life. Moreover, these associations may vary among subpopulations when stratified by factors that are known to influence sex hormone levels and mechanisms, such as use of hormone replacement therapy, menopausal status, and levels of obesity. To fill these critical research gaps, we propose to prospectively examine these associations with respect to risk of NHL [including the major subtypes: diffuse large B-cell lymphoma (DLBCL), follicular lymphoma (FL), and chronic lymphocytic leukemia (CLL)] and MM.

Given the previously described background and hypotheses, we will:

1. Determine the associations between (i) parity, (ii) age at first birth, (iii) age at menarche, (iv) age at menopause, (v) breastfeeding, and (vi) hormone replacement therapy use and risk of B-cell NHL, by major subtypes (DLBCL, FL, CLL), and MM. Analyses will be conducted for all women overall, and stratified by use of hormone replacement therapy, menopausal status, and BMI.

2. Determine the associations between BMI and history of diabetes and risk of B-cell NHL, by major subtypes (DLBCL, FL, CLL), and MM, overall and by gender.

3. Determine the meta-analytic summary associations for aims 1 and 2 when pooled results with the Women’s Health Initiative (WHI) cohort. To increase sample size and statistical power, we will have the opportunity to pool our results with those from our parallel proposal in the WHI cohort, which includes data from 478 DLBCL cases, 322 FL cases, 375 CLL cases, and 451 MM cases.