Return to Table of Contents
Female breast cancer incidence, mortality, and survival patterns differ by race/ethnicity (1,2). Based on invasive breast cancer cases diagnosed between 1986 and 1991, the National Cancer Institute's (NCI) Surveillance, Epidemiology, and End Results (SEER) Program showed the five-year relative survival to be 69% for black women, compared with 84.3% for white women (2). Similarly, Chapter 2 of this report has shown that while black women are less likely than non-Hispanic white women to develop breast cancer, they are more likely to die from it. Breast cancer survival is highly dependent on the stage at which the cancer was diagnosed. The five-year relative survival rate is 93% for cancers diagnosed while still confined to the breast, compared to 18% for those which have spread to distant parts of the body (2).
Several different coding systems exist for describing stage at diagnosis. The American Joint Committee on Cancer's (AJCC) classification of tumors uses information on tumor size, lymph node involvement, and spread to other organs, and is commonly used in clinical settings (3). A second system used to stage cancers at the time of diagnosis is referred to as the SEER Extent of Disease (EOD) system (4). This is a site-specific coding system that describes tumor size, tumor extension, and lymph node status. The SEER EOD coding system allows for collection of more detailed and specific tumor information than the AJCC staging system, and has been defined more consistently over time. EOD information can be translated into AJCC codes. A third system used for staging of cancers is known as the SEER Summary Stage system (5), which was utilized by the California Cancer Registry (CCR) during the time period covered in this report. In situ describes tumors with characteristics of malignancy but which have not penetrated the basement membrane of the tissue nor extended beyond the epithelium. A localized tumor is defined as one which is malignant and invasive but confined entirely to the organ of origin. Regional neoplasms have extended beyond the organ of origin into surrounding tissues, involve regional lymph nodes, or both. Finally distant tumors have spread to remote parts of the body from the primary site either by direct extension or by metastasis.
When attempting to compare stage at diagnosis of tumors that have been assigned stage based on the different staging systems, some difficulties arise. A breast tumor designated as in situ by SEER summary stage corresponds to Stage 0 based on the AJCC system, but a localized tumor may fall into either AJCC Stage I or II, depending on size of the tumor. A regional tumor of the breast may fall into AJCC categories II or III, again depending on tumor size, or if the tumor has extended to the chest wall or skin or has metastasized to axillary lymph nodes. Distant stage cancers are assigned to AJCC stages III or IV, depending on spread to secondary lymph nodes or distant metastases. The CCR has adopted the SEER EOD system for cases diagnosed in 1994 forward, enabling more direct comparisons with SEER and clinical criteria based on AJCC.
In the following chapter, data are presented on stage at diagnosis for female breast cancer cases diagnosed in California from 1988 to 1992. Stage at diagnosis categories used in the following analyses are based on the SEER summary stage, as described above. Stage at diagnosis is examined by race/ethnicity, age group at diagnosis, CCR reporting region, and socioeconomic status (education and income level of patient's surrounding neighborhood based on 1990 U.S. Census-defined block groups). Trends in stage at diagnosis from 1988 to 1992 are also analyzed by race/ethnicity, age group, and socioeconomic status. For selected analyses by age, age at diagnosis was grouped into cases diagnosed ages 0-49 (pre-menopausal) and 50 and older (post-menopausal).
Income and educational level associated with cancer cases reported in this chapter were characterized based on the level of the 1990 U.S. Census-defined block group surrounding the patient's home. A census block group was considered to have low income if its median household income was within the 25% lowest block group incomes in that CCR reporting region. High income block groups were those with median income within the 25% highest block group incomes in that region, medium income block groups were those with median incomes falling in the middle 50%. A patient was considered to live in a less educated block group if no more than 25% of people living in that block group had a high school diploma.
Statistical significance of temporal trends and possible associations between variables were evaluated with the chi-square test, with a p-value less than 0.05 considered significant.
From 1988 to 1992, 97,032 cases of in situ and invasive breast cancer were diagnosed in California women and reported to the CCR, excluding 31 cases of unknown age. The greatest proportion of diagnoses were staged as localized (53.1%), followed by regional (27.4%), in situ (11.8%), and remote (4.3%) (Table 7.1). Three and one-half percent of cases during that time period had unspecified stage at diagnosis.
The highest proportions of in situ breast cancer diagnoses were found among Asian/Other and non-Hispanic white women, 12.2% and 12.0%, respectively, followed by black (10.2%) and Hispanic (9.8%) women (Table 7.1, Fig. 7.1). Non-Hispanic white and Asian/Other women also had the highest proportions of localized diagnoses, 54.6% and 51.6%, respectively, while Hispanic women were diagnosed at the localized stage 46.2% and black women 44.6% of the time. Hispanic women had the highest proportion of regional diagnoses at 34.8%, black women the next highest at 33.5%, followed by Asian/Other women (29.4%). Non-Hispanic white women had the lowest proportion of regional breast cancer diagnoses, 26.1%, and the lowest proportion of remote diagnoses, 4.0%. Black women had the highest proportion of remote diagnoses, 7.5%, followed by Hispanic women with 5.1%. When comparing stage at diagnosis between race/ethnicities over the five year time period based on categorization of stages into early diagnoses (in situ or localized) and late diagnoses (regional or remote), a significantly greater percentage of cases was diagnosed at the early stage for both non-Hispanic white (66.6%) and Asian/other (63.8%) women than for either Hispanic (56.0%) or black (54.8%) women.
SEER data from 1992 revealed that the percent of cases diagnosed at the in situ stage was approximately equal for black and white women (13.4% and 13.3%, respectively) (2). In California in 1992, 13.1% of breast cancer cases in non-Hispanic white women reported to the CCR were diagnosed at the in situ stage, compared with 11.1% of breast cancer diagnoses in black women (data not shown). Data for non-Hispanic white women in California are not directly comparable to data from SEER for whites due to the fact that SEER includes both Hispanic and non-Hispanic white women in the category denoted white . When the stage distribution of invasive cancers diagnosed from 1986 to 1991 and reported to SEER was compared with 1988-1992 CCR data, it was found that California women tended to be diagnosed at an earlier stage. For non-Hispanic white women, 62.1% of invasive diagnoses in California were localized, 29.6% of diagnoses regional, and 4.6% classified as distant, compared with 59% localized, 32% regional, and 6% distant in the SEER data. Black women in California also tended to be diagnosed at an earlier stage compared with black women monitored by the SEER Program, 49.6% vs. 48.0%, respectively, diagnosed at the localized stage, 37.3% vs. 38.0%, respectively, at the regional stage, and 8.4% vs. 9.0% at the distant stage.
For women diagnosed with breast cancer between 1988 and 1992 who were less than 40 years old, in situ diagnoses comprised 10.3% of the total number of cases, localized 44.0%, regional 39.0%, and remote 3.8% of the diagnoses (Table 7.1, Fig. 7.2). In successively older age groups, the proportion of localized cases increased gradually, representing 45.5% of diagnoses in women 40-49, 50.4% of diagnoses ages 50-59, 56.6% of diagnoses ages 60-69, and 57.7% of diagnoses in women ages 70 and older. Also as age increased, the proportion of regional diagnoses decreased consistently. Localized and regional diagnoses together comprised approximately 80% of the total breast cancer diagnoses, regardless of age category. The consistent trend towards earlier diagnosis (greater proportion of localized and decreasing proportion of regional diagnoses) as age increased is probably, at least partially, due to the fact that, as age increases, breast tissue becomes less dense and tumors become more easily diagnosed (6). The proportion of in situ cases increased from 10.3% among women less than 40 to 16.0% in women ages 40-49, but then consistently decreased in women ages 50-59 (14.1%), 60-69 (11.4%), and 70 and older (8.9%). The proportion of remote diagnoses was fairly consistent across age groups, ranging between 3.7% in women ages 40-49 to 4.7% in women ages 70 and older.
When breast cancer stage distribution was analyzed by age (0-49, 50 and older) and race/ethnicity, significantly more cases were diagnosed at an early stage (in situ or localized) for women ages 50 and older than for women ages 0-49 for all four individual race/ethnicities (Fig. 7.3). The proportion of early stage diagnoses for Hispanic women increased proportionally by 15.5% when comparing women 50 and older to those less than 50, by 10.4% for non-Hispanic white women, by 6.9% for black women, and by 5.7% for Asian/Other women. Regardless of age category, non-Hispanic white and Asian/Other women were significantly more likely than non-Hispanic black or Hispanic women to be diagnosed at an early stage.
For all races combined, some variation between CCR regions was observed in stage at diagnosis for breast cancer cases diagnosed from 1988-1992 (Table 7.2). Regions 2 (Central Valley) and 9 (Los Angeles) had the lowest percentage of early diagnoses (in situ plus localized), 61.7% and 62.9%, respectively, while Regions 4 (Tri-County) and 8 (Bay Area) had the highest percentage of early diagnoses, 68.0% and 67.9%, respectively.
When comparing early versus late stage at diagnosis between CCR regions by race/ethnicity for 1988-1992, a greater amount of variation was noted than for all races combined, particularly for non-white women. For Asian/Other women, the proportion of early diagnoses ranged from 56.1% in Region 6 (North) to 71.4% in Region 4 (Tri-County). The percent of early diagnoses for black women ranged from 50.0% in Region 4 (Tri-County) to a high of 59.3% in Region 6 (North). The region with the lowest proportion of early breast cancer diagnoses in Hispanic women was Region 2 (Central Valley), 52.5%, and the region with the highest proportion was Region 8 (Bay Area), 62.7%. For non-Hispanic white women, the proportion of early diagnoses by region had a narrower range, from 63.3% in Region 2 (Central Valley) to 69.6% in Region 8 (Bay Area). Limited inferences should be drawn from these data due to the small numbers of cases diagnosed for non-whites in some of the regions (e.g., Regions 6 and 4). In all regions, the proportion of early diagnoses for black and Hispanic women was lower than the proportion for non-Hispanic whites. The lowest proportion of early diagnoses in white women (Region 2 (Central Valley), 63.3%) was higher than the highest proportion for black (Region 6 (North), 59.3%) or Hispanic (Region 8 (Bay Area), 62.7%) women.
The percent of female breast cancer cases diagnosed at the early stage from 1988 through 1992 by county is shown in Table 7.3. The percent of cases diagnosed at the early stage ranged from a low of 40.0% in Modoc county to a high of 71.5% in Marin county. These figures should be interpreted with extreme caution, however, because for a number of counties, e.g., Modoc, the overall number of breast cancer diagnoses was extremely small during that time. A graphical comparison of these county percentages to the statewide average (64.8% diagnosed at the early stage) is shown in Figure 7.4. Counties were designated to have percentages of early diagnoses equal to the statewide average if their respective percentages were within 1% of the state's. Those counties with percentages higher than the statewide average are located primarily along the coast of California, and at the northern end of the Central Valley. No attempt was made to verify if these differences were statistically significant because of the disparity in population sizes and number of cases.
For all race/ethnicities combined, the proportion of in situ and localized breast cancer diagnoses in California between 1988-1992 increased as the income level of the block group of residence increased (Table 7.4, Fig. 7.5). The proportion of in situ cases increased from 9.4% to 11.5% to 13.7% as income level changed from low to medium to high, respectively, and the percentage of localized diagnoses rose from 51.5% to 53.4% to 54.4%. Concomitantly, as the income level of the block group rose, the proportion of late stage diagnoses decreased. Regional diagnoses comprised 28.9% of cases from low-income block groups, 27.7% of cases from medium income block groups, and 26.2% of cases from high income block groups. Similarly, 5.9% of diagnoses from low income block groups were staged as remote, 4.3% from medium income block groups, and 3.4% from high income block groups.
The income level of the block group of residence at time of diagnosis was positively associated with the proportion of early breast cancer diagnoses for all four individual race/ethnicities (Fig. 7.6). The greatest relative increase in percent early stage at diagnosis was observed for Hispanic women, where the proportion increased from 52.9% to 63.7% (relative increase of 20.4%), from low to high income levels. The next greatest difference was seen in black women (relative increase of 9.6%), the next largest in Asian/Other women (relative increase of 7.3%), and the smallest difference was noted for non-Hispanic white women, where the proportion increased by 6.7% with higher income. Non-Hispanic white women diagnosed with breast cancer from low income block groups still were diagnosed at the early stage more often than Hispanic or black women from high income neighborhoods.
For all races combined, a positive association was observed between education level of a patient's block group of residence at time of breast cancer diagnosis and diagnosis at the in situ or localized stage (Table 7.5, Fig. 7.7). From 1988-1992, the proportion of in situ diagnoses in women from less educated block groups was 9.3%, compared with 12.7% of women residing in more educated block groups. Similarly, the proportion of localized cases was 50.8% in women from less educated block groups and 53.9% in women from more educated block groups. Correspondingly, the proportion of regional and remote diagnoses was higher in women living in less educated block groups than those living in more educated neighborhoods, 30.1% vs. 26.3% and 5.7% vs. 3.8%, respectively.
When the percent early stage at diagnosis was examined by race/ethnicity and education level of block group of residence, a higher percentage of early diagnoses was observed in women from more educated block groups for all four race/ethnicities (Fig. 7.8). Educational level appeared to have the greatest impact on stage at diagnosis for Hispanic women and the least effect for Asian/Other women. The greatest relative increase in the proportion of early diagnoses between the two education levels was observed in Hispanic women, 13.5%, followed by non-Hispanic white women, 6.8%, black women, 6.6%, and Asian/Other women, 1.7%. White and Asian/Other women from less educated neighborhoods were both diagnosed at the early stage a greater percentage of the time than more educated black or Hispanic women.
The annual number of breast cancer cases and the percent distribution of stage at diagnosis by race/ethnicity and age group (0-49, 50 and older) for California women diagnosed from 1988-1992 are shown in Table 7.6.
For women ages 0-49, all races combined, a slight but statistically significant upward trend in the percent of cases diagnosed at the in situ stage was observed, increasing from 13.8% in 1988 to 14.8% in 1992 (Fig. 7.9). A slight but statistically significant downward trend in the percent of diagnoses at the remote stage was noted, decreasing from 3.9% in 1988 to 3.3% in 1992. The proportion of localized and regional diagnoses varied less than the stages mentioned above, and the trends were not significant.
In women ages 50 and older, the percent of breast cancers diagnosed at the in situ and localized stage increased significantly from 1988 to 1992, and regional and remote diagnoses decreased significantly over this time period (Fig. 7.9). The proportion of in situ diagnoses increased from 9.9% in 1988 to 12.2% in 1992, and the proportion of localized diagnoses from 54.5% to 57.2%. The percentage of regional diagnoses declined from 26.8% in 1988 to 23.6% in 1992, and remote diagnoses declined from 4.8% to 4.1% over the same time period.
A significant increase, from 10.2% in 1988 to 14.7% in 1992, was observed in the trend for in situ breast cancer diagnoses of pre-menopausal Asian/Other women, but for no other stages for either age group of this race/ethnicity (Table 7.6). The distribution of stage at diagnosis for black women also did not change substantially between 1988 and 1992, except for a significant increase in the proportion of in situ diagnoses in women ages 50 and over, from 8.2% to 11.3%. For post-menopausal Hispanic women, there was a significant increase in the proportion of in situ diagnoses between 1988 and 1992 (7.3% to 10.7%), and a significant decrease in the proportion of regional diagnoses (33.5% to 28.6%). During this same time period, significant upward trends were observed for post-menopausal non-Hispanic white women for in situ (increasing from 10.1% to 12.3%) and local diagnoses (increasing from 55.7% to 58.3%), and significant downward trends for regional (decreasing from 25.9% to 22.8%) and remote (decreasing from 4.5% to 3.8%) diagnoses, as determined by chi-square analysis for trend.
When trends in percent early stage at diagnosis (in situ and localized combined) from 1988 to 1992 were examined by race/ethnicity and age (0-49, 50 and older), the proportion of early stage diagnoses was greater in 1992 than in 1988 for pre-menopausal black, Hispanic, and non-Hispanic white women, but none of the trends were significant (Fig. 7.10). Among women ages 50 and older, all four race/ethnic groups revealed significant upward trends. The percent early diagnoses increased for Asian/Other women from 63.2% to 68.7%, for black women from 51.7% to 61.0%, for Hispanic women from 56.6% to 62.6%, and for non-Hispanic white women from 65.8% to 70.6%.
Trends for early diagnosis of breast cancer for all races combined from 1988 to 1992 are shown by income level in Figure 7.11. The increase in the percent of early diagnoses for all three income groups is statistically significant. The proportion of low income women diagnosed at the early stage increased from 59.2% to 64.1%, the proportion of medium income women from 63.5% to 67.4%, and the proportion of high income women from 65.5% to 70.7%.
Trends for early stage at diagnosis by education level for all races combined from 1988 to 1992 are seen in Figure 7.12. Significant increases in the proportion of early diagnoses have occurred for women of both higher and lower education levels, but consistently more women living in more educated neighborhoods were diagnosed at an earlier stage. The proportion of early diagnoses for less educated women increased from 58.6% to 63.6% from 1988 to 1992, and for more educated women from 64.9% to 68.3%.
Between 1988 and 1992, 65% of all female breast cancer cases were diagnosed at an early stage (in situ or localized) of the disease. When analyzed by race/ethnicity, 67% of non-Hispanic white women, 64% of Asian/Other women, 56% of Hispanic women, and 55% of black women were diagnosed at an early stage. The disparity observed between black and non-Hispanic white women is comparable to that reported by SEER (2). When stage distribution was analyzed by age (0-49, 50 and older) and race/ethnicity, significantly more cases were diagnosed at an early stage for the post-menopausal women than for pre-menopausal women of all four individual race/ethnicities. Regardless of age, black and Hispanic women were more likely than non-Hispanic white or Asian/Other women to be diagnosed at a late stage (regional or remote).
Some variation in the percent of early stage at diagnosis was noted by CCR reporting region, but regardless of region, the basic race/ethnicity differences remained, i.e., non-Hispanic white and Asian/Other women had the highest proportion of early diagnoses and black and Hispanic women the lowest. A higher income and educational level was associated with an earlier stage of breast cancer diagnosis for all four race/ethnicities in California from 1988 to 1992.
Despite disparities in stage at diagnosis by race/ethnicity, income, and education, the percent of breast cancers diagnosed at the early stage has increased steadily over the five-year time period. When trends in the percent early stage (in situ and localized) at diagnosis from 1988-1992 were examined by race/ethnicity and age (0-49, 50 and older), significantly increasing trends were observed for post-menopausal women of all four race/ethnicities. Early stage diagnoses increased for Asian/Other women 50 and older from 63.2% in 1988 to 68.7% in 1992, for black women from 51.7% to 61%, for Hispanic women from 56.6% to 62.6%, and for non-Hispanic white women from 65.8% to 70.6%. Statistically significant increases in the proportion of early stage breast cancer diagnoses were seen for low, medium, and high income neighborhoods and from less and more educated neighborhoods.
Return to Table of Contents
Return to Chapter 6
Go to Table 7.1
Go to Chapter 8