updated September 25, 2015
Exposures to known and suspected risk factors for breast cancer begin early in fetal development and continue throughout life. Some relate to individual choices and lifestyle while others are encountered as an inevitable result of the way we design our communities and society more generally. In The Ecology of Breast Cancer I explored this multifactorial, multilevel complexity with an eye toward lessons learned from complex systems analysis while proposing practical interventions that may help prevent breast cancer and improve outcomes after diagnosis and treatment.
Risk factors are not experienced in isolation. They co-occur and interact in complex ways, creating system conditions in which breast cancer is more or less likely—in individuals and populations. When we attempt to take this complexity apart for purposes of research or decision-making, we often miss the properties that emerge from the whole.
A recent study may turn out to provide yet another example of these interactions at work. It has sparked new interest in the potential role of viral infections in the origin of breast cancer. But it’s the virus in combination with other environmental exposures where biologic plausibility of the hypothesis emerges and gains strength.
The authors studied the breast tissue of 114 women with breast cancer and 104 women without the disease. They report that breast cancer was three times more likely in women whose breast epithelial tissue contained DNA from a bovine leukemia virus (BLV), after adjusting for age, race/ancestry, and geographic area. BLV frequency in the breast tissue of an additional 21 women with premalignant changes—e.g. atypical ductal hyperplasia and carcinoma in situ—was intermediate between cancer cases and controls.
Another virus, the mouse mammary tumor virus (MMTV), has long been known to cause mammary tumors in mice, and controversy about its potential role or that of a similar virus in the origin of human breast cancer has ebbed and flowed for decades. The recent finding of an association with BLV reinforces the need to continue to look for viral infections that may contribute to breast cancer risk.
Cancer-causing viruses are implicated in at least six kinds of human cancer—hepatocellular carcinoma (hepatitis B and C viruses), cervical cancer (human papillomavirus [HPV]), Burkitt’s lymphoma and nasopharyngeal carcinoma (Epstein-Barr virus), adult T-cell leukemia (human T-cell leukemia virus [HTLV-1]), and Kaposi’s sarcoma (human herpes virus 8). MMTV-initiated tumor in the mouse is the only known animal breast cancer caused by a virus. MMTV is a retrovirus transmitted from mother to offspring via milk. Searches for a similar virus in human milk or breast tissues have had inconsistent results.
According to this study’s authors, bovine leukemia virus is a retrovirus closely related to HTLV-1 and causes bovine leucosis—a leukemia/lymphoma of B cells. Approximately 38% of beef herds, 84% of dairy herds, and 100% of large-scale dairy operation herds in the USA are infected with BLV. Fewer than 5% of these cattle develop clinical leucosis. BLV-infected lymphocytes circulate through the blood of infected cattle. BLV also infects the mammary epithelial cells of cows and may be present in cow’s milk although pasteurization renders BLV noninfectious.
Other cancer-causing retroviruses can integrate into host DNA, influencing the expression of various genes that can increase cancer risk. Deltaretroviruses, like BLV, do not have to integrate into DNA to have that effect. Proteins produced by HTLV-1 and BLV interfere with DNA repair and infected cells can then begin to accumulate DNA mutations caused by internal oxidative stress or exposures to other environmental agents. Thus, BLV infection alone may not be sufficient to cause breast cancer, but in combination with DNA-damaging agents the resulting genomic instability can ultimately lead to malignancy.
In animal herds, cow-cow transmission and biting flies can disseminate the virus. Animal husbandry practices can also spread the virus within a herd. But how humans become infected with BLV is not known. Transmission from cattle to humans is possible for people in close contact with the animals. If milk drinking is a pathway, and if pasteurization truly renders the virus non-infectious, then people consuming unpasteurized (raw) milk products would be at risk. Poorly cooked meat is another possibility. Human-human transmission is also plausible.
The authors mention several strengths of their study including biological plausibility, the fact that a viral origin of mammary gland cancer has been demonstrated in mice, an earlier demonstration that humans are infected with BLV (see Bovine leukemia virus DNA in human breast tissue), and the gradient of breast tissue diagnosis in relation to frequency of BLV exposure.
But there are also limitations. The validity of a case-control study like this depends on the assumption that cases and controls come from very similar populations. In this study, most case samples came from women who had mastectomies for breast cancer and most control samples came from reduction mammoplasties; they could represent different underlying populations.
Replication of the findings by others would help to validate them. And a prospective study showing that viral infection preceded cancer development would be essential to establish a causal association. BLV infection could also be more likely in circumstances that otherwise increase breast cancer risk, where the presence of the virus is simply a marker for that increased risk. Regardless, these provocative new findings point to a need for additional research into viral contributors to this all-too-common disease.