Return to Table of Contents
Although breast cancer is commonly referred to as a single entity, the location of the tumor within the breast and the type of tissue involved may have important etiologic, treatment and prognostic implications. This chapter presents a detailed description of the laterality, primary site and histology of female breast cancers diagnosed in California over the five-year period 1988-1992, as reported to the population-based California Cancer Registry (CCR)(1). The primary site and histology classifications used are those of the International Classification of Diseases for Oncology (ICD-O), Field Trial edition (2), which was used by the CCR from 1988-1991. Data collected in 1992 were coded using the second edition of ICD-O, and were converted to the Field Trial codes.
The ICD-O divides the breast into subsites so that the specific location of the tumor in the breast can be recorded. Vertical and horizontal axes through the nipple define quadrants: upper-inner (ICD-O-FT code 174.2), lower-inner (174.3), upper-outer (174.4) and lower-outer (174.5). Three subsites identify other specific portions of the breast: nipple and areola (174.0), central breast (directly under the nipple or areola, 174.1) and axillary tail (174.6). Subsite 174.8 includes tumors on the midline between quadrants and those described as inner, outer, upper or lower breast without assigning the tumor to a specific quadrant. When the medical record does not contain enough information to assign the tumor to one of these subsites, the primary site is recorded as "Breast, NOS (not otherwise specified)" (174.9).
The female breast is composed of 15 to 20 separate mammary glands, or lobules, which are connected by ducts to the nipple. The glands themselves are formed from a system of minute ducts and secretory cells, separated by connective tissue and surrounded by adipose tissue. As will be described below, about 95% of all breast cancers arise from the glandular epithelial lining of the ducts. However, coding procedures have been developed to distinguish the specific tissue of origin and clinical features of the tumor (2). A brief description of the most common histologic types of invasive breast cancer, focusing on the gross rather than microscopic features, is given in Table 5.1 (3,4).
During the time period covered in this chapter, the CCR used the National Cancer Institute's SEER (Surveillance, Epidemiology and End Results Program) summary stage guidelines to classify stage at diagnosis (5). In situ tumors have not infiltrated the basement membrane of the epithelium and therefore have not invaded the surrounding breast tissue. Localized tumors have invaded surrounding breast tissue, but have not spread beyond the breast itself into either lymph nodes or other adjacent organs. Regional tumors have extended from the breast into adjacent tissues and/or lymph nodes, and distant tumors have already spread to non-adjacent organs (metastasized) at the time of diagnosis. For a more detailed discussion of stage at diagnosis, please refer to Chapter 7 of this report (6).
A total of 97,032 breast cancers (in situ and invasive) were diagnosed among California women over the five-year period 1988-1992 (1). Of these, 48.5% occurred in the right breast, 50.4% in the left breast and 0.1% were bilateral; in 1.0% laterality was unknown. Cancers diagnosed in both breasts at the same time were coded as bilateral only if the histologic type was the same in both breasts and the point of origin was not specified. These figures therefore underestimate the frequency with which tumors are simultaneously identified in both breasts.
Excluding bilateral tumors and those of unknown laterality, 49.0% of tumors diagnosed in California occurred in the right breast and 51.0% occurred in the left, with a ratio of left to right tumors of 1.04 (Table 5.2). An excess of left-sided tumors has been consistently reported in the literature, with left/right ratios varying from 1.05 to 1.20 (7).
Left-sided tumors were more common at all stages of diagnosis, but the difference was more pronounced among tumors diagnosed at the in situ or distant stage (left/right ratios 1.06 and 1.10, respectively). Among in situ breast tumors, the left-sided predominance was more marked for post-menopausal women (50 years old or older, left/right ratio 1.07) than for pre-menopausal women (less than 50 years old, left/right ratio 1.04) (Table 5.2). Invasive tumors had the same degree of left-sided predominance in pre- and post-menopausal women (ratio 1.04 for both age groups).
When breast cancer laterality was examined by stage at diagnosis and race/ethnicity, an excess of right- rather than left-sided tumors was seen among Hispanic women with in situ tumors (ratio 0.93) and among black women with tumors diagnosed at the regional stage (ratio 0.99) (Table 5.2). All other stage- and race-specific comparisons showed left-sided predominance, with ratios varying from 1.01 to 1.20.
Laterality was also examined by location within the breast. Although left-sided tumors were more common than right-sided for most locations, in situ tumors of the central breast were more common in the right breast (left/right ratio 0.90) (Table 5.3). In addition, the most common location of invasive tumors, the upper-outer quadrant, showed an equal proportion of left- and right-sided tumors. To control for the simultaneous effects on laterality of race/ethnicity, histology and stage at diagnosis while retaining sufficient numbers for meaningful comparisons, cases were restricted to non-Hispanic white women with in situ or localized ductal carcinoma, NOS (ICD-O-FT histology code 8500) of the breast. In the restricted analysis, central breast tumors continued to show a right-sided excess among in situ tumors for both pre- and post-menopausal women, and also showed a right-sided excess for localized tumors among pre-menopausal women (Table 5.4). Upper-outer quadrant tumors showed a right-sided excess among localized tumors for both pre- and post-menopausal women, and for in situ tumors among pre-menopausal women.
A total of 85,596 invasive breast cancers were diagnosed among California women during the five-year period 1988-1992 (1). About 15% of these tumors had primary site reported as "Breast, NOS" (Table 5.5). This is considerably lower than the comparable figure (44.4%) reported by the National Cancer Institute's SEER (Surveillance, Epidemiology and End Results) Program for 1973-1977 (8), the most recently published data on detailed primary site from a population-based registry in the United States, reflecting improvements in the quality of cancer registration over the last twenty years.
Among the invasive female breast cancers in California for which a site other than "Breast, NOS" was given, 44% were located in the upper-outer quadrant of the breast, 11% in the upper-inner quadrant, 7% each in the lower-outer quadrant and central breast, 6% in the lower-inner quadrant, 1% each in the axillary tail and nipple/areola and 22% on the midline or inner/outer/upper/lower portion of the breast (Table 5.5). This distribution is very similar to that reported by SEER for 1973-1977 (8). When "Breast, NOS" cases were excluded from SEER data, 41% were located in the upper-outer quadrant of the breast, 11% in the upper-inner quadrant, 9% each in the lower-outer quadrant and central breast, 5% in the lower-inner quadrant, 2% in the nipple/areola, 1% in the axillary tail and 22% in midline and other positions.
The location of invasive breast cancers was very similar for pre- and post-menopausal women in California and for each of the race/ethnic groups (Table 5.5). Black women were somewhat more likely to have site reported as "Breast, NOS" (18.7% compared to 15.3% among non-Hispanic white women). When cases coded "Breast, NOS" were excluded, the percentage of invasive breast cancers located in the upper-outer quadrant was 44% for non-Hispanic white women and black women, 43% for Hispanic women and 42% for Asian/Other women.
The percent of female breast cancers reported as "Breast, NOS" increased with later stage at diagnosis: localized (10.9%), regional (13.7%), distant (36.7%) and unstaged (59.7%). Combining distant and unstaged tumors, nearly half (47.0%) were reported as "Breast, NOS" (Table 5.6). However, the location of distant/unstaged tumors for which a specific subsite was listed was similar to that for localized/regional tumors, but with a higher proportion of midline lesions (29.1% and 22.4%, respectively) and a lower proportion of upper-outer quadrant lesions (38.7% and 44.3%, respectively). This reflects a tendency for distant/unstaged tumors to be reported with less specific information, rather than an increased risk for late-stage diagnosis among midline tumors.
A larger percentage of in situ breast cancers were coded "Breast, NOS" compared to invasive tumors (21.5% and 14.9%, respectively) (Table 5.5). The percentage of in situ breast tumors with an unspecified location was slightly higher among pre-menopausal women (23.2%) and Asian/Other women (23.9%), and was lowest among black women (16.5%). When cases coded "Breast, NOS" were excluded, the location of in situ breast cancers was very similar to that described above for invasive tumors: upper-outer quadrant (44%), upper-inner and central breast (8% each), lower-outer and lower-inner (7% each), nipple/areola (2%), axillary tail (less than 1%) and midline and other tumors (23%) (Table 5.6). When cases coded "Breast, NOS" were excluded, the percentage of in situ breast cancers located in the upper-outer quadrant was 45% for non-Hispanic white women, 44% for Hispanic women, 42% for black women and 38% for Asian/Other women.
Of the 85,596 invasive breast cancers diagnosed among women in California from 1988-1992, 84,176 (98.3%) were microscopically confirmed and were included in the following analyses. More than 75 distinct histologic types of invasive breast cancer were reported (Table 5.7). However, 71.6% were a single histologic type, infiltrating duct carcinoma, NOS (ICD-O-FT histology code 8500). This was very similar to the comparable figure (72.5%) reported by SEER for 1983-1987 (9), the most recent population-based data available on breast cancer histology in the United States.
Aggregating some of the more similar histologic types, lobular carcinoma was the next most frequently diagnosed invasive breast cancer in California (7.2%), followed by infiltrating duct and lobular carcinoma (4.5%), comedocarcinoma (2.7%), adenocarcinoma, NOS (2.6%) and mucinous adenocarcinoma and mucin-producing adenocarcinoma (2.4%) (Table 5.8). Medullary, inflammatory and unspecified carcinomas and tubular adenocarcinomas each accounted for between one and two percent of invasive breast cancers.
The percentage of infiltrating duct carcinoma, NOS, was 69.3%, 70.9%, 71.8% and 73.5% among black, Hispanic, non-Hispanic white and Asian/Other female breast cancer cases, respectively. Race/ethnic differences were especially apparent among the less common histologies. Lobular and duct and lobular carcinomas and tubular adenocarcinomas were more common among non-Hispanic white women, while comedocarcinomas and mucinous adenocarcinomas were more common among Asian/Other women (Table 5.8 and Figure 5.1). Black and Hispanic women had a higher percentage of medullary and inflammatory carcinomas and of adenocarcinoma, NOS, than Asian/Other and non-Hispanic white women.
Although infiltrating duct carcinoma, NOS, was the most common histology among both pre- and post-menopausal women diagnosed with breast cancer (72.5% and 71.4%, respectively), the frequency of some of the less common histologies varied with age (Table 5.9 and Table 5.10). Comedocarcinoma and medullary carcinoma were relatively more common among pre-menopausal women in all four race/ethnic groups, and inflammatory carcinoma was relatively more common among younger women in all race/ethnic groups except Asian/Others. Lobular carcinoma and mucinous, tubular and papillary adenocarcinomas were relatively more common among post-menopausal women in all four race/ethnic groups. Figure 5.2 shows the proportions of selected histologic types over five-year age categories among non-Hispanic white women, for whom there were sufficient numbers of cases to examine changes with age in more detail.
Figure 5.3 compares the histologies of invasive female breast cancers diagnosed in California during 1988-1992 to those reported by SEER for 1983-1987 among 56,312 white and 4,585 black cases (9). Data for non-Hispanic whites in California are not strictly comparable to data from SEER for whites, since SEER included both Hispanic and non-Hispanic white women in the category "white." Both non-Hispanic white and black women with breast cancer in California had a lower percentage of non-specific histologic codes (adenocarcinoma, NOS, and carcinoma, NOS), and were two to three times more likely to have inflammatory carcinoma than white or black breast cancer cases from SEER. In addition, both non-Hispanic white and black women with breast cancer in California were more likely to be diagnosed with lobular carcinoma, duct and lobular carcinoma, comedocarcinoma, mucinous adenocarcinoma and tubular adenocarcinoma than their counterparts in SEER. Because more recent data from SEER were not available at the time of the publication of this report, it could not be determined if these differences reflect changes in the relative frequency of histologic types over time or between geographic areas, or both. SEER has reported temporal increases in the proportion of breast cancers which were inflammatory carcinomas and lobular carcinomas, and decreases in the proportion of less specific histologies (11). In addition, it is unclear whether differences that do exist reflect different diagnostic patterns among physicians or true differences in the risk of developing certain types of breast cancers.
Table 5.11 shows the location of invasive breast cancers with a specific location reported (174.0-174.5) for the ten most common specific histologic types. In general, the location of tumors was similar for most histologies: roughly 60% of tumors with a specific site occurred in the upper-outer quadrant, 15% in the upper-inner quadrant, 10% in the lower-outer quadrant, 10% in the nipple, areola and central breast combined, and 5% in the lower-inner quadrant. The major exceptions to this pattern were Paget's disease, which predominantly occurred in the nipple, areola and central breast (64%), and inflammatory carcinoma, which also had a relatively high proportion (29%) of tumors in the nipple, areola and central breast. Mucinous and papillary adenocarcinomas also occurred somewhat less frequently in the upper-outer quadrant (47% and 48%, respectively).
Although the location of invasive breast cancers was very similar among pre- and post-menopausal women, as discussed above, some differences between these two groups were seen for specific histologic types (Table 5.11). A higher proportion of Paget's disease occurred in the nipple, areola and central breast among women aged 50 and older (67.4%) than among younger women (52.0%). Similarly, a higher proportion of cases occurred in the nipple, areola and central breast of older women than of younger women with inflammatory carcinoma of the breast (33.4% and 21.6%, respectively) and with medullary carcinoma (8.7% and 3.9%, respectively).
Table 5.12 shows the location of invasive breast cancers for the six most common histologic types by race/ethnicity. Although the number of race- and histology-specific cases was sometimes small for black, Hispanic and Asian/Other women, the same general pattern described above for all races combined was seen for each of the race/ethnic groups. However, Asian/Other women had a higher proportion of ductal carcinomas, lobular carcinomas and duct and lobular carcinomas located in the inner breast compared to non-Hispanic white women.
Certain histologic types are more likely to be diagnosed at the earlier, localized, stage than others (Table 5.13). Examining invasive breast cancers which were microscopically confirmed and staged, the percent of cases which were localized was highest for tubular adenocarcinomas (93.3%), followed by mucinous and mucin-producing adenocarcinomas (87.5%), papillary adenocarcinomas (84.3%), comedocarcinomas (72.6%) and medullary carcinomas (67.8%). The most common histologic types (duct, lobular and duct and lobular) had almost the same percentage of cases diagnosed while localized (63.0%, 62.8% and 61.9%, respectively). Despite the small number of cases among black, Hispanic and Asian/Other women for histologic types other than duct carcinoma, a very similar pattern is seen for all four race/ethnic groups.
All 11,436 in situ female breast cancers diagnosed in California from 1988 to 1992 were microscopically confirmed. More than twenty histologic types were identified (Table 5.14). As with invasive breast cancers, intraductal carcinoma, NOS (ICD-O-FT 8500) was the most common histologic type, but accounted for a smaller proportion of cases, 55.2%, compared to 71.6% of invasive cases. SEER reported that 60.9% of the 2,350 in situ female breast cancers diagnosed from 1983 to 1987 were intraductal carcinoma, NOS (9).
The percent of in situ breast cancers in California which were ductal carcinoma, NOS, was 54.3%, 54.8%, 60.1% and 60.4% for black, Asian/Other, Hispanic and non-Hispanic white women, respectively (Table 5.15). Comedocarcinoma was the second most common in situ cancer for each of the four race/ethnic groups, accounting for 18% of in situ cancers overall, and, as seen for invasive breast cancers, was more common among Asian/Other women (Figure 5.4). Lobular carcinoma was the third most common in situ breast cancer among non-Hispanic white, black and Hispanic women and was the fourth most common among Asian/Other women. Papillary carcinoma was the third most common in situ breast cancer for Asian/Other women, and the fourth most common for the other three race/ethnic groups. Together, comedocarcinoma, lobular carcinoma and papillary adenocarcinomas accounted for 37.1% of all in situ breast cancers, compared to 10.5% of invasive breast cancers.
Lobular carcinoma in situ was about two times more common among pre-menopausal than post-menopausal women in California among all four race/ethnic groups (Table 5.10). As with invasive breast cancers, the proportion of papillary carcinomas increased with age in all race/ethnic groups. Among Asian/Other women, comedocarcinoma in situ was more common among pre-menopausal women, but was more common among post-menopausal women in the other race/ethnic groups.
Comparing race-specific data from California in 1988-1992 to data from SEER in 1983-1987 (9), non-Hispanic white and black women with in situ breast cancer in California had a lower proportion of lobular carcinomas, although the difference is more dramatic for non-Hispanic white women (Figure 5.5). Comedocarcinoma represented a considerably higher percentage of in situ breast cancers for both race/ethnic groups in California, accounting for 18.3% and 13.1% of non-Hispanic white and black cases, respectively, in California, and 5.0% and 7.7% of white and black cases, respectively, from SEER. As discussed above, there is not sufficient information to completely interpret these differences.
As has been consistently reported elsewhere, data from California show that breast cancer is somewhat more likely to occur in the left breast than in the right. The reasons for this are not clear, but it has been shown that in about 55% of women, the left breast is slightly larger than the right (12), and therefore contains a larger amount of tissue at risk for becoming cancerous. Some relatively common exceptions to left-sided predominance are noted in this chapter.
Female breast cancers occur most often in the upper-outer quadrant, regardless of race/ethnicity, age at diagnosis (pre- or post-menopausal) or stage at diagnosis, and for most of the common histologic types of breast cancer.
Although there are more than 75 histologic types of breast cancer, the vast majority (71.6% of invasive cancers and 55.2% of in situ tumors) are ductal carcinomas, not otherwise specified. Differences in histologic type among race/ethnic groups and by age are apparent among the less common histologies. Compared to non-Hispanic white women, Asian/Other women have a higher proportion of comedocarcinomas and mucinous adenocarcinomas and black and Hispanic women have a higher proportion of medullary and inflammatory carcinomas, whether pre- or post-menopausal. The proportion of breast cancers which are comedocarcinomas or medullary carcinomas decreases with age, while the proportion of mucinous or papillary carcinomas increases with age.
Some of the less common histologic types of breast cancer (e.g., tubular, mucinous and papillary adenocarcinomas and comedocarcinoma) are more likely to be diagnosed at the earlier, localized stage. The CCR does not have sufficiently complete follow-up information on women diagnosed with breast cancer to calculate survival rates at this time. However, based on data from the SEER program, these histologic types also have a higher five-year relative survival rate compared to infiltrating duct, lobular or duct and lobular carcinomas (11).
Return to Table of Contents
Return to Chapter 4
Go to Table 5.1
Go to Chapter 6