Mouse Virus Evidence Suggests Viral Basis for Breast Ca

SAN ANTONIO—A large proportion of human breast cancers may be associated with the human mammary tumor virus (HMTV), which is nearly identical to the murine mammary tumor virus (MMTV) that is implicated in breast cancer in mice. HMTV is transmissible in vitro and can infect human mammary epithelium along with cultured B and T lymphocytes. Together, these findings suggest that breast cancer may have an infectious etiology, according to investigators from Mount Sinai School of Medicine, New York, who have for years been pursuing this theory. Principal investigator James F. Holland, MD, Distinguished Professor of Neoplastic Diseases at Mount Sinai, presented updated findings from his work with Dr. Beatriz Pogo and colleagues at the 29th Annual San Antonio Breast Cancer Symposium (abstract 6).

According to Dr. Holland, the transmissibility of HMTV offers a possible explanation for the worldwide differences in the incidence of breast cancer. In support of his theory, he pointed out that about 15% of human cancers are associated with viruses—for example, cervical cancer, hepatomas, and lymphomas. "We thought that if human breast cancer were due to a virus, the virus would be a 'kissing cousin' of the murine mammary tumor virus," Dr. Holland said, explaining in an interview that HMTV could be a mutation from the mouse virus.

The search for MMTV revealed a 660-base pair (bp) segment (env) that is completely unique from sequences found in all other viruses. Sequencing of the human provirus revealed a 95% sequence homology to the mouse virus, with most differences representing frame-shift mutations in the long-terminal-repeat region typically associated with endogenous retroviruses.

Using this gene segment—"a footprint of the viral presence"—they detected HMTV in 30% to 40% of human breast cancer samples provided from centers in the Americas, Europe, and Australia, and in 60% to 80% of samples from Africa (excluding Kenya). In contrast, only 0% to 12% of samples from China, Japan, and Vietnam (reported by Ford) contained HMTV.

Furthermore, they found that normal tissue does not contain HMTV. The virus was detected in 30% of breast cancer samples from the NCI tissue registry, but in less than 1% of normal tissue from the same breast; only 1.4% of samples from mammoplasty and 0% of tumors other than breast cancer contain HMTV.

The fact that HMTV env is undetectable in normal breast tissue from persons with HMTV-positive tumors indicates that the virus is not genetically inherited, Dr. Holland said.

Interestingly, the wide geographic variation in the HMTV env gene sequence detection rate corresponds to the natural habitat of Mus domesticus, the common house mouse. Human breast cancer incidence rates are notably higher in Western Europe, roughly coinciding with areas in which Mus domesticus is dominant, compared to Eastern Europe, where a different mouse species, Mus musculus, is most common.

To Dr. Holland, this strongly supports the notion of an infectious transmission. "We have lots of theories about the mode of transmission, but we have no proof," he told reporters. "We have to establish that women get infected before they get breast cancer."

In vitro experiments have indeed shown that HMTV can be transmitted "horizontally" through infection. "Characteristic B type retroviruses bud from primary breast cancer cell cultures, with molecular properties identical to those of the provirus," he said. "In co-culture experiments, HMTV can be transmitted through a viral filter to recipient human mammary epithelia and to cultured human B and T cells."

He noted that there are anecdotes of human infection with MMTV prior to breast cancer, which support the plausibility of HMTV as a causal agent for human breast cancer. "These findings raise the possibility that a major segment of human breast cancer is caused by an infection, explaining in part the worldwide differences in incidence," he said.