Skip to main content
         

Vol. 05 Issue 3, Early Fall 2000

Research Commentary: Breast Cancer Incidence Highest in the Range of One Species of House Mouse
The Ribbon 

Stewart, T.H.M., R.D. Sage, A.F.R. Stewart and D.W. Cameron. British Journal of Cancer 82 (2000):446-45.

This study compared the incidence of human breast cancer to the geographical areas inhabited by different types of mice. Human breast cancer incidence varies to a large extent between different geographical areas. Mice living in different geographical areas may have different percentages of their populations infected with the mouse mammary tumor virus (MMTV). This virus causes mammary tumors in mice and these investigators propose that it can undergo several changes to allow it to cause breast cancer in humans. They also propose that the presence of MMTV infected mice explains the geographical differences in the incidence of breast cancer.

It is important to realize that this study was designed only to generate a hypothesis to be evaluated by subsequent investigations. The idea that certain types of mice give humans cancer was not examined directly but rather the investigators used data from other studies to put forth this hypothesis. In epidemiology, studies of this type are called ecological studies. Since they are designed to generate hypotheses, they are only required to set forth a possible explanation. No proof is expected or required.

The authors do not present a strong hypothesis. It is built on unestablished information and remotely possible phenomena. Central to this hypothesis is the idea that there are areas occupied by mice highly infected with MMTV and areas occupied by mice lowly infected by this virus. There is no foundation for such a proposal, as studies have not been conducted to directly determine locations occupied by mice with potentially active virus. Two types of mice do exist which seldom live in overlapping areas in various regions of the world. Most dramatically, there is a division in the living range of these mice between the western and eastern parts of Europe. Although both these types of mice can carry the MMTV gene, studies have shown that they can differ in the number of copies of the MMTV gene they do carry. However, the mice were from too few locations for the authors to make predictions about the presence of MMTV in mice in other areas. For example, the authors focus on differences in infected mice in Europe but the presence of MMTV in one of the types of mice has never been examined anywhere in Europe. The authors assume that North American mice, which have been examined, would be infected to the same extent as those in Europe because the North American mice are thought to have originated in Europe. Geographical localization of mice wasnot the focus of these studies and they were not designed to answer this question reliably. The authors have made worldwide predictions about the presence of MMTV in mice from studies of a few locations. The extent of infection of the different types of mice at different geographical locations is uncertain and should not be used as the foundation for a hypothesis.

MMTV is a retrovirus and during infection adds its gene to the genes of the host mouse. Mice can carry a number of copies of this gene. How active MMTV becomes depends not on the number of copies the animal carries but rather where the MMTV gene is inserted, that is, what genes it has as its neighbors. The studies, which looked at the presence of MMTV in the two mouse populations, found that about 85% of one type of mice carry MMTV, whereas, 40% of the other type of mice carry MMTV. This is a two-fold difference and does support the author's idea that one type of mice would be more likely to be infected. However, the authors were not satisfied with this size of difference. They devised their own criteria, based on the number of copies of MMTV the mice carried, to artificially increase the infection difference of the two types of mice. Since the mice that had the gene less frequently also had fewer copies, this division based on copy number dramatically changed the comparison of the mice. There was no biologically proven basis for this division and it was misuse of the data.

MMTV, like all members of its class of viruses, is able to undergo recombination. This means that MMTV could potentially pick up from other viruses the ability to infect humans. Such a change would be very rare but is possible. A recombination allowing MMTV infection of other species has not been observed, no other species exhibits breast cancer with MMTV-like biology. Most importantly, breast cancer in humans behaves differently from MMTV-induced cancer in mice. The spread of MMTV between mice has only been demonstrated to occur through breast milk. Infected mothers produce virus in their breast milk, which is transferred to nursing pups. Virus in the breast milk infects the pup's white blood cells and breast epithelial cells. The potential for breast feeding as a route for human breast cancer has been examined directly by a large case-control study involving 8300 women with breast cancer (Titus-Ernstoff et al., 1998). This study found that transmission of breast cancer through breast milk was unlikely (see discussion in accompanying article by SR Ross). Differences also exist in the tissues infected. Unlike MMTV infected mice, traces of the virus are not found in normal breast tissue and white blood cells of women with breast cancer who do have MMTV-like DNA in their breast tumors (Wang et al., 1995). The behavior of the virus during pregnancy in mice also differs from human breast cancer. MMTV induced mammary tumors in mice are increased by pregnancy whereas, pregnancy decreases the risk of breast cancer in humans. MMTV would thus have had toundergo changes in its route of transmission, tissues infected and its activation by pregnancy to behave in a manner compatible with human breast cancer biology. Such major changes in the behavior of the virus would be very unlikely.

The authors finally separate first Europe and then the rest of the world based on the ranges of the two types of mice. They perform a statistical analysis of human breast cancer incidence in the European areas, which coincides with east and west Europe. They find a significant difference in breast cancer incidence between the two regions. There is also an area in Europe where both types of mice live. Here, they find intermediate breast cancer rates. Breast cancer incidence in other parts of the world was also presented relative to their mouse populations. The results were shown to agree somewhat with the area's mouse distributions. The authors attribute these differences to the presence or absence of MMTV infected mice but other well established explanations exist. The dissimilarity, between these regions, which is accepted by most epidemiologists, is that of affluence. Women living in Western Europe have a much higher standard of living than women in Eastern Europe. Breast cancer has been associated with affluence through the independent development of a number of risk factors. Breast cancer risk is higher in women who: do not have children; have children at a later age; have more education or income; do not breastfeed; have early menarche or late menopause, and; have a taller stature. Affluence impacts all of these factors. It should, nonetheless, be kept in mind that these risk factors and other risk factors could explain at best half the cases of breast cancer. Hypotheses, such as this one, serve the important functions of pointing to areas where knowledge is incomplete and stimulating thought and discussion.

Presentation of this paper in the popular press gave the impression that mice had been demonstrated to be possible carriers of breast cancer. It was not made clear that this was a hypothesis put forth by one group of investigators in an ecological study. One can identify ecological studies by the fact that the investigators are comparing an environmental factor to the incidence of some disease and producing a hypothesis.

References

Titus-Ernstoff L., K.M. Eagan, P.A. Newcomb , J.A. Baron, M. Stampfer, E.R. Greenberg, B.F. Cole, J. Ding, W.M. Willett, D. Trichopoulous. "Exposure to milk in infancy and adult breast cancer risk." J Natl Cancer Inst 90 (1998):921-924.

Wang, Y., J.F. Holland, I.R. Bleiweiss, S. Melana, X. Liu, I. Pelisson, A. Cantarella, K. Stellrecht, S. Mani, and B.G.T. Pogo. "Detection of mammary tumor virus ENV gene-like sequences in human breast cancer." Cancer Res. 35 (1995):5173-5179.

Barbour S. Warren is a Research Associate for the BCERF program;
he received his Ph.D. in Pharmacology and Toxicology from the University of Louisville Medical School.

Back to the top

Last Update 08.23.00