Aspirin and Breast Cancer Prevention: Title and subTitle BreakThe Estrogen Connection

In this issue of THE JOURNAL, Terry et al¹ report that aspirin use in women is associated with a significant reduction in the risk of breast cancer, especially for hormone receptor–positive tumors. Previous studies have examined the relationship between breast cancer risk and aspirin or other nonsteroidal anti-inflammatory drug (NSAID) use but have not delineated different subtypes of breast cancers to determine this kind of correlation. This report is the first to examine whether the protective effect of aspirin varies with estrogen receptor (ER) or progesterone receptor (PR) status. With the genomic and proteomic revolution well underway, it is becoming quite clear that there are distinct differences in molecular signatures between tumors derived from the same organ. It is well known that the hormonal responsiveness of breast cancers is important for determining the treatment regimen and can affect the clinical outcome. Thus, it is not surprising that one subset of breast cancers may be more responsive to a particular prevention strategy than another.

More than 80 million aspirin tablets are consumed each day in the United States.² This 100-year-old drug is one of the most commercially successful pharmaceutical agents ever produced. Yet despite the widespread use of aspirin over several decades, one of the mechanisms of its action was not really known until the early 1970s when Vane³ found that it blocked the production of "proinflammatory" prostaglandins. Based on this discovery, Vane and 2 Swedish scientists, Bergstrom and Samuelsson, won the Nobel Prize in Physiology or Medicine in 1982.⁴ Later it was found that aspirin binds covalently and blocks the active site of the prostaglandin-endoperoxide synthase 1 enzyme (often referred to as cyclooxygenase 1 [COX-1]).⁵ A second, inducible form of prostaglandin-endoperoxide synthase (COX-2), which is also affected by aspirin, was subsequently cloned and the complementary DNA sequence was reported in 1991 by at least 3 groups.^{6 - 8}

Aspirin use has been associated with a reduction in mortality from cardiovascular disease⁹ and colorectal cancer.¹⁰ Observational studies reporting a protective effect of aspirin on breast cancer have been mixed. Liu et al¹¹ addressed the question experimentally by creating COX-2 transgenic mice in which expression was driven by the murine mammary tumor virus (MMTV) promoter. Breast carcinomas developed spontaneously in multiparous female MMTV–COX-2 mice, indicating that COX-2 alone has oncogenic potential in the mammary gland. Chang et al¹² later delineated the molecular mechanism(s) by which COX-2–derived prostaglandin E₂ (PGE₂) induces tumor-associated angiogenesis, which is required for the initiation and/or progression of mammary cancer in MMTV–COX-2 mice. These investigators reported that PGE₂ induced angiogenesis at the earliest stage of tumor development, even before PGE₂-induced mammary gland hyperplasia. They also found that the nonselective NSAID indomethacin inhibited both PGE₂-induced angiogenesis and breast tumor progression and confirmed the role of COX-2 by using the COX-2 selective inhibitor celecoxib. This work clearly demonstrated that COX-2 overexpression can cause breast cancer in mice.

The current report by Terry et al finds that the inverse association between aspirin use and breast cancer was clearly evident for every patient subgroup except for those with negative hormone receptor status (ER−PR−). The association was strongest among frequent aspirin users. However, acetaminophen use was not associated with protection in any subgroup. These findings are noteworthy and may reveal some important mechanistic insight about a connection between aspirin and estrogen. In 1996, a report by Subbaramaiah et al¹³ suggested transcriptional activation of COX-2 in transformed mammary epithelial cells. That same year, Zhao et al¹⁴ demonstrated that PGE₂ can induce aromatase expression leading to increased estrogen production in mammary adipose stromal cells. In 1999, Soslow et al¹⁵ found that COX-2 was expressed in breast ductal carcinoma in situ and in breast cancers. Recently, other investigators reported that COX-2 is up-regulated in the normal adjacent epithelium to ductal carcinoma in situ and that COX-2 overexpression coincides with focal areas of p16INK4a hypermethylation in vivo that could represent early neoplastic changes leading to breast cancer.^{16 - 17}

Taken collectively, these studies provide a clear rationale for a role of COX and prostaglandins in breast cancer. Blocking COX activity with aspirin or other NSAIDs would inhibit aromatase induction and result in lower estrogen levels (Figure 1). Therefore, the observation that receptor-positive tumors are more responsive to aspirin is consistent with these preclinical and clinical observations. This association needs to be confirmed before clinicians can make any definite recommendations to patients at risk for breast cancer. However, it does appear that there is emerging evidence supporting a protective effect of aspirin in ER+ and PR+ breast cancers.

Despite the longstanding and ubiquitous nature of aspirin use, researchers are still exploring the clinical outcome of aspirin treatment in humans. Unfortunately, all the answers are not available and current information is insufficient to make any definite recommendations to patients. Women who take daily aspirin for cardiovascular indications may gain additional benefits with regard to reduction in their risk for certain cancers, such as hormone receptor−positive breast cancer. However, the optimal aspirin dose or regimen required to achieve a maximal reduction in cancer risk remains unknown.