Discussion
Several major international congresses have been held optimizing breast cancer care among women from several countries, and all of them have noted East-West differences [
56]. The goal of the Breast Surgery International symposium attended by surgeons from many nations during the 2007 Montreal International Surgical Week was to discern whether breast cancer characteristics other than incidence differ between Asian and Western countries. The summaries presented by breast cancer experts from China, Taiwan, India, Japan, South Korea, Sweden, Canada, and the Unite States confirm previous studies in showing a striking difference in peak incidence, which is between 40 and 50 in Asian countries, and between 60 and 70 in Western countries. The relative contribution of biological, genetic, and environmental factors to this difference clearly merits further study if we are to determine whether breast cancer is the same disease in Asia as in the West. Incidence is increasing in Asia and in the West, but mortality is rising in Asian countries and decreasing in the West.
Female breast cancer in the Western world has shown a steady yearly increase of 1.5%. The most commonly cited reasons for this increase are breast cancer awareness, early detection by screening programs, and reliable diagnostic tools. Screening on a national basis exists in several countries, namely Canada, the United Kingdom, the Netherlands, Finland, Sweden, Norway, Australia, and Singapore. Preventive measures, and effective surgical, endocrine, cytotoxic, and radiation therapy account for the excellent survival data and the slight decline in mortality during the last decade in the West. Although these countries do not represent the world as a whole, the data collected give us a glimpse of the breast cancer statistics in these countries and are a critical first step toward broader global representation in the future.
The importance of a global approach to breast cancer is to share the preventive measures and appropriate treatments from more advanced countries rapidly with developing countries [
1,
3,
56]. Although the infrastructure for reporting cancer varies widely in the countries that participated in the symposium, and the data often are not representative of the entire country, important global comparisons can still be made. The trend shown in this report is that the incidence of breast cancer increases as countries become more urbanized. The traditional approach of mastectomy gives way to more lumpectomies. Although sentinel lymph node biopsy is a standard staging procedure [
51‐
53] for early breast cancer in the U.S., and perhaps in Canada and Sweden, it is still a new technique and being evaluated in Asia.
To attempt to answer the question posed by the title of this article—is breast cancer the same disease in Asia as in Western countries?—we can first take an overview of the data presented at the ISW 2007 meeting. The differences in the data obtained are striking. If we put to one side data on the stage at presentation of cancer (which is most likely related to access to service, and will be a major determinant of survival) and look instead at incidence rates, it is clear that incidence rates for breast cancer are uniformly lower in Asia than they are in the West.
National cancer registry data have confirmed that breast cancer incidence rates differ between Asia and the West in two major ways: first, the overall rates are far lower [
57,
58]; second, the shape of the age-incidence curve is markedly different and importantly, this difference mirrors differences in the age-incidence distribution for ER+ and ER− breast cancers [
59]. When breast cancer incidence is plotted as a log-log plot, the linear increase in breast cancer is inflected around the menopause (“Clemmesen’s hook”) [
60]. When one plots ER+ and ER− breast cancers separately, the slopes are quite distinct, with the ER+ risk continuing after the menopause, whereas the ER− clearly diminishes [
61]. This is also seen when breast cancer histological subtypes are plotted independently; for example, medullary breast cancers (which are very frequently ER−), show a sharp falling-off in incidence after the menopause [
62]. There is also evidence for this bimodality in risk profile when intrinsic breast cancer subtypes [
63], rather than ER status, are used to subdivided breast cancers. The median age at diagnosis for luminal A and B tumors was 74 years, whereas it was 52 years for basal and HER2+ subtypes [
62], suggesting that bimodality is an essential feature of breast cancer worldwide, and that the two main differences between breast cancer in the West and in Asia is the prevalence of the relevant risk factors and the access to health care.
It is possible that some of the difference in risk profiles between Western and Asian women is related to the structure and gene expression profile of the normal breast. For example, normal breast epithelium is much more likely to be ER+ in Western women than in Japanese women [
64]. By contrast, the data on breast density is conflicting, and in general, Asian women in Asia, compared with whites in North America, seem to have low breast cancer rates despite having small, dense breasts, although it may be that within Asian populations, risk is proportional to density [
65]. The relationship between density and risk is also uncertain when studying Asians born and/or living in North America, as the data are conflicting [
66‐
68], and it is not clear if correcting for hormonal factors and body mass index can completely explain any differences observed.
Other more general factors, such as total energy balance, are likely to be important as well, and may interact with breast-specific effects as well as nonspecific hormonal “drivers,” to modulate the risk for breast cancer. This leads on to consider what happens when environments change but genes stay the same.
When families move from Asia to the West, they take their genes with them, but to a large extent, they leave their environment behind. While, naturally, some aspects of home life will be similar for new immigrants, they will inevitably be exposed to the new world they find themselves in. Data clearly show that when the female offspring of Asia-born couples are brought up in North America, their risk for breast cancer is substantially increased compared with the rates observed in Asia-born Asian females [
69]. Nevertheless, incidence rates among Asia-born Asian females are clearly increasing at a rapid pace [
57,
58,
70,
71]. The only plausible explanation for these findings is that some aspect of what is variously called urbanization, “modern living,” and “Westernization” is affecting Asian breast cancer rates quite rapidly and probably in an irreversible way. This process seems to be affecting different populations in slightly different ways.
The second way to study this phenomenon is to study a population in which the environment is roughly the same but the genes are different. Singapore is a city, an island, and a state and is composed of three main ethnic groups: Chinese, Malays, and Indians. In the past 40 years, Singapore has experienced a rapid economic growth. From 1968 to 2002, 85% of 16,178 female breast cancers in Singapore occurred in Chinese, 10% in Malays, and 5% in Indians. Over this time period, breast cancer has increased year by year; by 3.1, 2.8, and 1.7% per year for Chinese, Malays, and Indians, respectively. Despite the rather similar external environments, current breast cancer rates are somewhat different in the three ethnic groups, and they have changed in different ways over the time period studied. Essentially, what is seen is that up to age 50, rates are lowest in Indians, but rise in a similar fashion in the other two groups. After age 50, the pattern is different; the rates for Indians continue to rise to age 70 years, whereas for Chinese, and even more so for Malays, the rates decrease. Birth cohort effects are also quite marked, especially in the Malays, where women born in 1971–1975 have seven times the risk of breast cancer as those born in 1921–1925 [
71]. Interestingly, all three populations experienced very similar and dramatic declines in fertility, but the temporal trends in breast cancer incidence differ between the groups, suggesting that other factors are playing a role in determining breast cancer risk.
As discussed above, one can specifically analyze subtypes of breast cancer. Ideally, one would like to be able to study many different breast cancer subtypes at the same time in different ethnic and geographical subgroups (e.g., basal-like, luminal-like, etc.) but this is impractical at present. It is possible, however, to study estrogen receptor-positive (ER+) and estrogen receptor negative (ER−) subtypes, as a large proportion of laboratories capable of making a diagnosis of breast cancer will be able to offer basic immunohistochemical analysis for the presence or absence of the estrogen receptor. When one compares ER+ and ER− breast cancer separately, some very interesting findings emerge. First, when considering Western countries, it is clear that ER+ and ER− breast cancers are different diseases [
61].
The proportion of ER+ to ER− breast cancers does seem to differ somewhat by North American ethnic groups, with blacks having the highest percentage of breast cancers being of the ER−/PR− phenotype (23%), followed by Hispanics (18.1%) and Filipinos (17.4), with Japanese and whites having almost identical percentages (14.1 and 14.7%, respectively) Therefore, other than blacks compared to all others, variations in ER status by ethnicity do not seem to be large [
72]. Interestingly, the age peak for ER−/PR− tumors occurs much later in life for Japanese Americans and native Hawaiians (64–65 years) than for whites (50 years) [
72]. Detailed data are generally not available for Asian countries.
One might expect that worldwide differences in incidence rates would be greater for ER+ cancers than they are for ER− cancers, but it is difficult to prove this as women in developing countries will have less access to laboratory and imaging technologies, and this could lead to various biases that would diminish the number of ER+ breast cancers diagnosed [
73]. Nevertheless, a detailed analysis of the literature has revealed significant differences in the effect of known epidemiological risk factors (such as parity, age at first birth, and body mass index) on ER+ compared with ER− breast cancer [
74,
75]. In general, ER− cancers seem somewhat insensitive to most of the traditional risk factors associated with breast cancer, making prediction and prevention of such cancers rather difficult. Conversely, since ER+ cancers are sensitive to “lifestyle” factors, one would predict that is it ER+ cancers that are going to increase rapidly in numbers in Asians as they become more Westernized [
76].
Taken together, these findings suggest the following seven conclusions:
-
Rates of breast cancer are uniformly many times lower in Asia than in the West
-
The peak age of onset in Asia is 45–50 years of age, whereas it is 55–60 years in the West
-
There is a clear deficiency of post-menopausal ER+ breast cancers in Asians living in Asia compared with Western populations
-
The decrease in age-specific breast cancer rates with age seen in Asian women cannot be completely explained by a deficiency of ER+/PR+ cases
-
Outside of some special situations, there is no biological difference in the breast cancers occurring in Asia and the West
-
The observed differences are most likely mainly attributable to different risk factors acting differentially on two (or more) different types of breast cancer
-
As Asia becomes urbanized and Westernized, we can expect significant changes in the currently observed patterns, towards Western-type data.
In response to these conclusions, the participants in the Breast Surgery International symposium made a recommendation to establish an international prospective database, which could perhaps be funded by the World Health Organization. Such a database would be a more reliable way to determine similarities and differences in breast cancer among women from diverse countries and different ethnic and genetic backgrounds, as the data will be prospectively collected.