Question ID: WS-81
Submitted by: Thomas Mack
March 30, 2011
A PROVOCATIVE QUESTION DOES THIS OBSERVATION OUT OF THE BOX WARRANT ATTENTION? An ounce of prevention should still be worth a pound of cure. However, while improvements have been made in the treatment of breast cancer, the number of cases actually prevented is still woefully small, despite the existence of strong predictors of high risk such as age, obesity, family history, and mammographic density. Even the net number of lives saved by secondary prevention, i.e. mammographic screening, is limited[1]. In contrast to the high risk Mendelian genetic predictors evident in linkage studies, responsible for a small minority of heritable cases, the genetic determinants of most heritable cases, as identified to date by GWAS studies, are very diverse, each with limited predictive value [2]. Even the predictive value of family history is diminishing, given the decreasing average American family size and the strong influence of lifestyle. Moreover, those lifestyle factors have proven difficult to manipulate. It seems unlikely that science will enable better means of prevention in the foreseeable future. Additional GWAS studies will probably not enable the screening of individual women, although they may slowly permit a better understanding of pathogenesis. Seven years ago, an observation was published that, if true, gives reason to consider another strategy for the prevention of heritable breast cancer[3]. A subgroup of women was found to be at a level of risk much higher than that of the average American woman. This subgroup is sizeable and growing, and an increasingly large number of women worldwide will become members. The observation is consistent with the body of past literature, withstood the highly skeptical reviewers of a prestigious journal, and was accompanied by a laudatory editorial[4]. Neither the observation, the methodology, nor the interpretation has ever been challenged. Nonetheless, the implications have not been exploited. This may be because the conclusion was (and is) “out of the box”, not in accord with conventional wisdom Specifically, 1811 pairs of female twins affected by breast cancer were identified, divided on the basis of zygosity and discordant/concordant breast cancer, and the differences between co-twins investigated in respect to standard breast cancer risk factors. It had previously been shown that unlike the fraternal co-twin of a breast cancer case, whose risk approximates that of a non-twin sibling, the risk to an identical co-twin of a case is much higher, at a level in excess of four times baseline[5], meaning that concordantly affected North American identical twins represent a genome at very high risk. The comparison of each high risk fraternal twin case to her unaffected co-twin produced a somewhat muted replication of the standard comparison of singleton cases to controls, i.e. a modest increase in risk after early menarche, delayed menopause, nulliparity, and a delayed first full-term delivery, all characteristics consistent with a carcinogenic result of high cumulative exposure to endogenous ovarian hormones, slightly confounded by genome. Because the age-specific incidence of breast cancer increases rapidly with increasing age until about menopause, then adds a relatively constant increment of 2-3% annually, it was reasoned that cumulative protection from breast cancer is, in practice, largely a function of age and competing mortality. Indeed, if the age-specific incidence curve for white California residents in 2008 is projected to older age groups, essentially all women would become affected by the age of about 130, and, aside from the small number of women at risk from Mendelian genes such as BRCA1/2, the phenotype of the more common polygenic forms of breast cancer can be thought of as reflecting a generic downward shift in age-specific incidence ( a fact clearly evident among the siblings and co-twins of breast cancer cases[6]). It therefore seemed reasonable to identify the determinants not only of breast cancer, but also of earlier diagnosis, among women matched on high genetic risk. In each of the 209 identical twin pairs in which both twins had been diagnosed with breast cancer, we compared the twin with the earlier diagnosis to the twin with the later diagnosis, considering the same standard risk factors. We found one very strong predictor of earlier onset of heritable breast cancer, namely age at puberty, whether identified by age at menarche, the age at onset of regular cycles (both long recognized as risk factors among unselected singleton cases), or especially by the age at first breast appearance (thelarche, a reliable comparative[7], especially memorable among twin girls). Puberty coming earlier than that of the co-twin was 5.3 times more likely to be followed by a breast cancer diagnosis diagnosed earlier than that of the co-twin, a highly significant finding, even though most of the breast cancers among these reproductively unexceptional women were post-menopausal. Most dramatically, if the earlier puberty appeared before age 12, the relative risk of having an earlier breast cancer diagnosis shot up to 9.1. These findings were interpreted to suggest that this age-specific increase in incidence resulted not from a cumulative excess in hormonal exposure, but a high heritable susceptibility on the part of immature breast ductal cells to hormone exposure. Almost all (more than chance would predict) of these concordant paired cases were concordant for hormone receptor positivity (5). It was reasoned that the subsequent interim uncontrolled growth of cancer cells occurs very slowly, probably only during the short progestin exposure within the luteal phase of each cycle. The observation is consistent with the available literature, i.e. not only the usual magnitude of early menarche as a standard risk factor among singleton populations[8], and the strong environmentally-driven secular increase in risk among BRCA1 carriers [9] (the latter in tandem with the widespread drop in average age at puberty) [10], but the known susceptibility of the immature breast to the carcinogenic effects of ionizing radiation[11]. Events have given cause for more concern. A secular drop in age at female puberty over recent decades has occurred on all continents[12-23], including Africa[24]. This decrease has recently slowed within developed populations and the affluent members of developing populations [25], but the trend toward younger age, even among women in developed countries, is now clearly stronger for thelarche than for menarche[26-29], and the earlier the breasts appear, the longer the interval between thelarche and menarche[30]. This change cannot be explained fully by changes in socioeconomic or nutritional status, including the emerging obesity epidemic[26], and it is not possible to dismiss the possibility that exposure to endocrine disrupting chemicals is partly responsible. Unfortunately, among this increasingly large proportion of girls with early thelarche, there is sure to be a subgroup with high heritable risk, and among them breast cancer is likely to become much more frequent. If the early breast development in the face of a family history of breast cancer does predict high lifetime risk, intervention is called for, but the means are not obvious. The available means of delaying menarche with GnRH agonists or gonadotropins[31] is crude and not target-organ specific, despite recent progress in understanding the interaction between CNS peptides, the hypothalamus, the pituitary, and the ovary[32]. Targeted support for follow-up research is needed. Others should be encouraged to replicate the original finding, and we plan to do so, although enough affected twin subjects are hard to accumulate. Attempts to reduce the toll of heritable breast cancer should certainly go beyond the narrow goal of identifying additional high risk polymorphic loci. Even if the observation cannot be replicated (and especially if it can), there is abundant reason to assign priority to the following questions: 1. Do the healthy immature breast ductal cells of girls at high heritable risk differ from those of others in the response to ovarian hormones (and endocrine-disrupting chemicals)? 2. As age at thelarche decreases, do healthy breast ductal cells (especially those of girls at high heritable risk) differ in cytology, growth pattern, or the response to ovarian hormones (and endocrine-disrupting chemicals)? 3. Is age at puberty determined only by genetics and stature, but also by particular nutritional, physiological, or toxicological determinants? 4. Are age at thelarche and age at menarche separately determined by cumulative nutritional, physiological, or toxicological determinants? 5. Can interventional strategies be designed to separately delay menarche and thelarche? Breast cancer will more frequently mar the adult lives of girls at heritable risk in all populations as more nascent breasts are exposed to a sudden flood of ovarian hormone. It is not unreasonable to hope that this burden of will someday be substantially reduced by collective or individual interventions, hopefully without the need for detailed reference to the many pertinent genotypes. Thomas Mack M.D., M.P.H. Ann Hamilton Ph.D. 1. USPSTF, Screening for Breast Cancer: USPSTF Recommendation Statement. Ann Int Med, 2009. 151: p. 716-26. 2. Stratton, M. and N. Rahman, The emerging landscape of breast cancer susceptibility. Nat Genet, 2008. 40: p. 17-22. 3. Hamilton, A.S. and T.M. Mack, Puberty and genetic susceptibility to breast cancer in a case-control study in twins.[comment]. New England Journal of Medicine., 2003. 348(23): p. 2313-22. 4. Hartge, P., Genes, hormones, and pathways to breast cancer.[comment]. New England Journal of Medicine., 2003. 348(23): p. 2352-4. 5. Mack, T., et al., Heritable breast cancer in twins. British Journal of Cancer, 2002. 87: p. 294-300. 6. Peto, J. and T. Mack, High constant incidence in twins and other relatives of women with breast cancer. Nature Genetics, 2000. 26: p. 411-4. 7. Berg-Kelly, K. and I. Erdes, Self-assessment of sexual maturity by mid-adolescents based on a global question. Acta paediatr, 1997. 86: p. 10-17. 8. Hsieh, C.-C., et al., Age at menarche, age at menopause, height, and obesity as risk factors for breast cancer: associations and interactions in an international case-control study. Cancer, 1990. 46: p. 796-800. 9. Kotsopoulos, J., et al., Age at menarche and the risk of breast cancer in BRCA1 and BRCA2 mutation carriers. Cancer Causes and Control, 2005. 16: p. 667-74. 10. Levy-Lahad, E. and S. Plon, A risky business--assessing breast cancer risk. Science, 2003. 302: p. 574. 11. Miller, A., et al., Mortality from breast cancer after irradiation dirung fluoroscopic examinations in patients being treated for tuberculosis. N Engl J Med, 1989. 321: p. 1285-9. 12. Rubin, C., et al., Timing of maturation and predictors of menarche in girls enrolled in a contemporary British cohort. Paed Perinat Epidemiol, 2009. 23: p. 492-504. 13. Chavarro, A., et al., Socio-demographic predictors of age at menarche in a group of Columbian University women. Ann Human Biol, 2004. 31: p. 245-57. 14. Rigon, F., et al., Update on age at menarchein Italy: toward the leveling off of the secular trend. J Adolesc Health, 2010. 46: p. 238-44. 15. Cabanes, A., et al., Decline in age at menarche among Spanish women born from 1925 to 1962. BMC Pub Healrth, 2009. 9: p. 449. 16. Aksglaede, L., et al., Recent decline in age at breast development: the Copenhagen Puberty Study. Pediatrics, 2009. 123(e932-9). 17. Tomova, A., et al., Menarche in Bulgaria-secular trend in the twentiewth century. Akusherstvo I Ginekologia, 2009. 48: p. 10-4. 18. Kalichman, L., et al., Age at menarche in a Chuvashian rural population. Ann Human Biol, 2008. 33: p. 290-7. 19. Semiz, S., et al., Pubertal development of Turkish children. J Pediat Endocrin, 2008. 21: p. 951-61. 20. Babay, Z., et al., Age at menarche and the reproducive performance of Saudi women. Ann Saudi Med, 2004. 24: p. 354-6. 21. Rah, J., et al., Age of onset, nutritional determinants, and seasonal variation in menarche in rural Bangladesh. J Health, Population and Nutrition, 2009. 7: p. 802-7. 22. Cho, G., et al., Age at menarche in a Korean population: secular trends and influencing factors. Europ J Pediatr, 2010. 169: p. 89-94. 23. Wang, D. and M. Murphy, Trends and differentials in menarcheal age in China. J Biosoc Sci, 2002. 34: p. 349-61. 24. Umeora, O. and V. Egwuatu, Age at menardche and the menstrual pattern of Igbo women of southeast Nigeria. `African J Reproduct Health, 2008. 12(`): p. 90-5. 25. Euling, S., et al., Examination of US puberty-timing data from 1940 to 1994 for secular trends: panel findings. Pediatrics, 2008. 121 S-3: p. S172-S191. 26. Aksglaede, L., et al., Age at puberty and the emerging obesity epidemic. PloS one, 2009. 4: p. e8450. 27. Codner, E., et al., Age of Pubertal events in Chilean school-age girls, and its relationship with socioeconomic status and body mass index. Revista Medica de Chile, 2004. 132: p. 801-808. 28. Ma, H., et al., Onset of breast and pubic hair development and menses in urban Chinese girls. Pediatrics, 2009. 124: p. e269-e277. 29. Herman-Giddens, M.E., et al., Secondary sexual characteristics and menses in young girls seen in office practice: a study from the Pediatric Research in Office Settings network [see comments]. Pediatrics, 1997. 99(4): p. 505-12. 30. Marti-Henneberg, C. and B. Vizmanos, The duration of puberty in girls is related to the timing of its onset. Ann Human Biol, 1997. 24: p. 61-4. 31. Boepple, P.A., et al., Use of a potent long-acting agonist of gonadotropin-releasing hormone in the treatment of precocious puberty. Endocrine Reviews, 1986. 16: p. 198-206. 32. Tena-Sempere, M., Kisspepin signaling in the brain: recent developments and future challenges. Molecular and Cellular Endocrinology, 2010. 314: p. 164-9.
| This question has not yet been evaluated by users |
Comments
