Breast Cancer: 20-Year Comparative Mortality and Survival Analysis by Age, Sex, Race/Ethnicity, Stage, Grade, Disease Duration, Selected ICD-O-3 Oncophenotypes, and Cohort Entry Time-Period: A Systematic Review of 656,501 Cases for Diagnosis Years 1975-2019: (SEER*Stat 8.4.0.1)

In breast cancer patients, total overall 5-year relative survival in 2 time-periods 1998-2019 and 2012-2019, is 88.1%, and 92.9%, respectively, all ages, stages and grades combined. In the 1998-2019 period, White female relative cumulative survival-SR was 92.9% with MR 174% at 0-5 years, and Black female SR was 83.1% with MR 327%, 47% >than Whites (to be discussed in the Commentary). During the 1998-2019 time-period, using SEER historic stage A (Tables 1-9), low grade-local stage breast cancer displayed the highest 5-year relative survival at slightly greater than 101% (Table 1) with median observed and relative survival >than 20 years. High grade, distant stage breast cancer, AJCC Stage IV, (Tables 10 & 11) manifested the lowest 5-year relative survival, 26.8%; with median observed and relative survival, 2.2 and 2.4 years, respectively (SEER adjusted AJCC 6^th ed. Staging).

Breast cancer is a molecularly diverse disease categorized into 3 major biological subtypes based on the presence or absence of phenotypical molecular biomarkers for estrogen or progesterone receptors and human epidermal growth factor 2 (ERBB2; formerly HER2): hormone receptor positive/ERBB2 (negative in >70% of patients), ERBB2 positive (15%-20%), and triple-negative (tumors lacking all 3 molecular markers, 15%). Appraising receptor status is critical to guiding therapeutic decision-making. Current approaches and evolving strategies for local and systemic therapy have been recently reviewed in 2019.¹⁷  Acquired risk factors for breast cancer are well known in their association with estrogen.¹⁸  However, and especially provocative, are risk factors with a genetic/hereditary predisposition, which significantly influence screening and follow-up recommendations for high-risk women. In patients with a suggestive personal and/or family history, a specific predisposing gene is identified in <30% of cases. Up to 25% of hereditary cases are due to a mutation in one of the few identified rare, but highly penetrant genes (BRCA1, BRCA2, PTEN, TP53, CDH1, and STK11), which confer up to an 80% lifetime risk of breast cancer. An additional 2%-3% of cases are due to a mutation in a rare, moderate-penetrance gene (eg, CHEK2, BRIP1, ATM, and PALB2), each associated with a twofold increase in risk.^19,20  In an individual without cancer, the immune system attacks cells with abnormal DNA or growth. This mechanism fails in those with breast cancer leading to tumor growth and spread. However, the genetic influence and metabolic basis for breast cancer disease remain topics in evolution, and thus are not discussed further here.

The National Cancer Institute SEER program with its government-issued, deidentified, public-use database since 1973 remains an excellent resource because its cancer information has remained consistent over time and does not require institutional review board approval based on guidelines from the National Human Research Protections Advisory Committee.

In females the 2015-2019 rate of 132.9/100,000 is the highest incidence for all female cancer sites. In males the rate is 1.2/100,000 and only 0.9% of that for females. In the 2015-19 time-period, the rate per 100,000 increases with age from a minimum of 1.5 at ages 20-24 to 133 at 40-44, to a maximum of 464.2 at 75-79, then decreases to 327.1 at age 85 and up. By race the 2015-19 rate is highest in non-Hispanic whites (135.5); in blacks (128.3) and lowest in American Indians, Alaska natives and Asian/Pacific Islanders (113.7). The overall percent change (PC) trend in whites, 1975-2019, was 27.8% and the slower rate annual percent change (APC) trend in whites was 0.4% from 1975 to 2019. In the same time-period, the overall percent change (PC) in blacks was 30.0% with 0.7% APC.

As of January 1, 2019, the Estimated Prevalence Count was 676,498 (4.03%) in a population of 13,721,745 people. Populations were estimated by averaging 2018 and 2019 populations (SEER*Stat8 8.4.0.1).

SEER Case Statistics. Cases were excluded if there was no FU (diagnosis by autopsy or death certificate) or if there was a second primary cancer.

Cancer Case Statistics: Frequency distributions by age, sex, race, stage, grade, and ICD-O-3 histologic onchophenotypes are displayed in Chart 1:

• Total number of breast cancer patient-entrants, male & female combined, 1975-2019 was 656,501 cases.

• By sex: female 651,972 (99.3%), male 4529 (0.7%).

• By race: 1975-2019, White 549,925 (83.8%), Black 46,017 (7.0%), Other (American Indian/Alaska Native, Asian/Pacific Islander) 58,339 (8.9%).

• Ages <35 2.1%, ages 35-64 53.7%, ages 65 up 44.2%.

• The apex of frequency by diagnosis in both females and males was 65-69 years.

• Mean ages by sex, in males and females, 1975-2019, was 66.8 and 61.9 years, respectively.

• Mean ages by race, in Whites, Blacks and Other, 62.6, 57.4 and 59.0 years, respectively.

• Mean ages for selected ICD-O-3 histologic cell types for males and females are exhibited.

• All stages (known & unstaged) by race, 1975-2019, male and female combined;

• Total staged & unstaged cases, all races combined 553,729 and 16,800, respectively; race, ethnicity, both sexes combined,

• In the 1998-2019 cohort, 6,349 cases or 2.0% were unstaged,

• All races, 63% were local stage, and 37% were combined regional & distant stages,

• Whites, 64% were local stage, and 37% were combined regional & distant stages,

• Blacks, 54% were local stage, and 46% were combined regional & distant stages,

• Other, 65% were local stage, and 35% were combined regional & distant stages,

• (Breast cancer staging disparities with corresponding mortality & survival outcomes by race and ethnicity, are discussed in the Commentary section of this article).

• All grades (known & unknown) by race, 1975-2019, male and female combined;

• In the 1998-2019 cohort, 28,441 or 8.0% were of unknown grade.

• Whites, 47% were grades 1&2 combined, 26% were grades 3&4 combined,

• Blacks, 39% were grades 1&2 combined, 39% were grades 3&4 combined,

• Other, 54% were grades 1&2 combined. 29% were grades 3&4 combined.

Case Characteristics.Chart A indicates that after exclusion of cases with no follow-up (FU), the total number of cases for which mortality and survival data were available per cohort entry period was reduced to 538,959 (82.1%), with about the same percentage male-female sex distributions as in the SEER frequency database. Comparative mortality and survival analysis for these cases in Tables 1-11 are given the 2 subdivided cohorts, 1975-97 and 1998-19, with groups organized by age, sex, race, stage and grade, with mean annual duration intervals to 20-years. Only 2% of all cases in the later cohort were unstaged, and 8% were of unknown grade.

Chart A, Figure 1: As seen in Chart A and Figure 1, only approximately 2% of all cases, both sexes combined, were diagnosed prior to age 35 years. More than 55% of cases were diagnosed at the time of breast cancer presentation below age 64. The crest of diagnostic frequency in males (14.9%) and females (12.2%) occurred at ages 65-69 years and then declined to ages 85+. The SEER historic coding system was used in Tables 1-9, and AJCC staging was used for the 1998-2019 cohort in Tables 10 and 11. The total number of female cases available for survival and mortality analysis, 1975-2019, was reduced from 656,501 to 538,959 (82.1%) of which 215,354 were in the 1975-1997 cohort and 319,997 in the 1998-2019 cohort because of the progressive increase in incidence rates and trends until 2019 as reflected by the increase in SEER registries. Female mean age was 62.0 years in the 1975-1997 cohort and 61.9 years in the 1998-2019 cohort. Compared with other anatomic sites, 1975-2019, breast cancer, both sexes & all races combined show most patients in the localized stage (63.0%). However, in Whites, both sexes combined, local stage 63.5%, regional stage 30.5%, distant stage 6.0%; Blacks, both sexes combined, local stage 54.1%, regional stage 36.0%, distant stage 9.9%; Other races, both sexes combined, local stage 64.8%, regional stage 29.5%, distant stage 5.7%. These racial staging disparities are discussed in the Commentary section.

Figure 2: The apex of frequency for Whites-occurred at age 65-69 years (12.5%), then declined, leveling-out at ages 85+ years. In Blacks and Others, the apices of frequency occurred at ages 50-54 years (12.5%), then declined, leveling out at ages 85+ years.

Figure 3: Female breast cancer cumulative relative survival-SR in the entire 1975-2019 investigative period, all ages, stages and grades combined, showed a steady decline from the initial 0-1 year interval of diagnosis, 97.8%, to terminal interval of 15-20 follow-up years, 73.1%.

Figure 4: Female breast cancer cumulative observed and expected mortality rates, q & q’, 1975-2019, all ages, stages and grades combined, are displayed in Figure 4. Observed mortality rate q in the 0-1 interval 0.0399, an annual rate, is 70% higher than the initial expected mortality rate q’ of 0.0183 with corresponding SR 97.8%. At the 2^nd, 1-2 year interval of disease-duration, observed q has risen to 0.0458, 13% progressively higher than at initial diagnosis with corresponding SR declining to 95.1%. From intervals 5-10 to 15-20 years, observed mean q begins a slow ascent from 0.0406 to 0.0451 with corresponding decline in SR from 81.4 to 73.1% (see Commentary for further considerations). (Fig. 1-4).

Detailed comparative mortality results are given in Tables 1-11 for a variety of combinations of cohort, stage, grade, age, duration and race/ethnicity. All tables contain aggregate mean annual mortality rates (q), with corresponding observed and excess deaths (d & d’), excess mortality as MR and EDR, and, cumulative relative survival as SR. Mean mortality rates q, (∑d/∑E) are weighted by the frequency of occurrence of death (d) in a defined population at risk of death (E) during a specified interval, thus as the number of deaths d increase in a specific time relative to exposure E, observed mortality rate q increases. Quinquennial case frequency statistics for sex and race/ethnicity to ages 85+ years are shown in Table A (Breast Cancer Case Statistics) and Figures 1 and 2. Follow-up interval duration mortality and survival data to 20-years are displayed in Figures 3 and 4. The range of these indices is very wide. In Tables 1-5, each cohort was classified for mortality and survival analysis by a combination of age, stage & grade into seven group intervals of aggregate average annual outcome data to 20-year follow-up in two subdivided cohort investigative time-periods, 1975-1997 and 1998-2019. In Table 6, results are juxtaposed, line by line, for the two cohorts, all ages combined, by each combination of stage and grade in the preceding Tables, with additional results for the ungraded cases. Mortality by sex and race/ethnicity are shown in Tables 7 and 8, respectively. In Tables 1-8, the separation of the experience into 2 cohorts permits comparison of excess mortality between the earlier and later subdivided phases of the observation period from 1975 through 2019. It also permits the results in Tables 10 & 11 by the AJCC staging system, not used for SEER coding until 1988.

Comprehensive frequency and mean ages by sex and race are shown in Chart A. In Table 9, survival and mortality of selected ICDO-3 histopathologic oncophenotypes are also analyzed, tabulated and displayed for the 1998-2019 time period. Overall loss to follow-up was very low at 0.5% at 5 years. Losses to FU did vary with age. Overall, mean annual mortality rates q, and excess death rates-EDR (q-q’ per 1000) increased with advancing age but were variable or increased with escalating disease duration, Mortality ratios, MR, (100d/d’) diminished with advancing age because expected deaths d’ increase with aging. In local stage, grades 1&2 breast cancer – best case scenario – for which MR is below 100%, comparative mortality is low and survival high. Mortality increases & survival diminishes with advancing stages & grades of cancer.

Observed and expected interval mortality rates (q & q’) in 146,065 female entrants with breast cancer, local stage, grades 1&2, (well or moderately differentiated), all ages combined, available for mortality and survival analysis in the SEER 1975-2019 database are displayed in 3 age groups, <65, >65 and all ages combined to the terminal 15-20 years disease-duration interval. At durational intervals from diagnosis of 0-5 to 15-20 years in the 1975-2019 observation period, MR ranged from 90 to a maximum of 113% with corresponding EDRs of -2.2/1000 at 0-5 and 5.4/1000 at 15-20 years with cumulative relative survival SR of 94.5% in the last interval. A curious pattern existed in the 1^st two intervals and also at 0-5 years with observed q less than expected q’. At 0-1 year, observed mean q 0.0123 and expected mean q’ 0.0194 with corresponding MR 63%, EDR -7.1/1000. Similar patterns are exhibited in both the early (1975-1997 and later (1998-2019) investigative cohorts. Observed mean mortality rate increased to its maximum from 5 to 20 years climbing from 0.0195 to 0.0455 with corresponding modest increases in excess mortality, MR 113%, EDR 5.4/1000; SR declined to 94.5%. Maximum relative cumulative survival, SR 100.7% occurred in the 1^st duration, then declined to 94.5% at 20 years. Mortality rates, q & q’, weighted by d ÷ E, increased with advancing age in both investigative cohorts.

1975-1997, 0-1 year:

• <age 65, observed mean-q 0.0037; expected mean-q’ 0.0050,

• >age 65, observed mean-q 0.0281; expected mean-q’ 0.0399.

1998-2019, 0-1 year:

• <age 65, observed mean-q 0.0034; expected mean-q’ 0.0045,

• >age 65, observed mean-q 0.0208; expected mean-q’ 0.0358.

Median survival times:

• 1975-1997: observed 18.3 years, relative >20 years,

• 1998-2019: observed >20 years, relative >20 years,

• 1975-2019: observed >20 years, relative >20 years.

(Fig 5)

1975-1997, All Ages Combined; there were 37,407 female patient entrants. Relative cumulative survival rate (SR=100P/P’) at 2-5 years disease duration is 99.5%, then declined to 90.7% at 20 years. A curious pattern existed within one year of diagnosis, all ages combined with observed mortality rate q 0.0157 less than expected q’ 0.0222, corresponding MR 71%, EDR -6.5/1000, and at 1-2 years disease-duration MR of only 95%, EDR -1.1/1000. Maximum MR of 118% did not occur until 5-10 years duration. Initial observed mean mortality of 0.0157 rose to 0.0457 in the terminal interval. Excess mortality persists to 15-20 years with MR 115%, EDR 6.2/1000 and SR 90.7% based on 4,655 interval deaths. Median observed & relative survival was 18.3 and >20 years respectively.

Ages <65 and >65 years; 2-5 years disease duration: MR shows the usual pattern of decrease with advancing age, from 206% at age <65 to 102% at ages 65 up. Observed mean q is 0.0130 but increases to 0.0496 at ages 65 up indicating a 74% increase in observed mean q with advancing age in 1st investigative cohort.

Median survival times:

• <65: observed is >20 years, relative is >20 years,

• >65: observed is 11.6 years, relative is >20 years.

1998-2019, All Ages Combined; there were 108,658 female patient entrants all with relatively favorable prognosis. Relative cumulative survival rate SR 101.6%, at 2-5 years disease duration then declining to 96.5% at 20 years. A similar curious pattern as seen in the 1975-1997 investigative cohort existed within 1-2 years of diagnosis, all ages combined, with observed mean mortality rate q 0.0151 less than expected q’ 0.0197, corresponding MR 77%, EDR -4.6/1000, SR 101.2%. and, Maximum observed mean mortality rate q, 0.0426, did not occur until the terminal duration 15-20 years with corresponding MR 111%, EDR 4.1/1000 and SR 96.5% based on 2,971 interval deaths Median observed and relative survival times were >20 years.

Ages <65 and >65 years; 2-5 years: MR shows the usual pattern of decrease with advancing age, from 120% at age <65 to 89% at ages 65 up. Observed mean q is 0.0067 but increases to 0.0389 at ages 65 up indicating an 83% increase in observed mean q with advancing age in the 2^nd investigative cohort.

Median survival times:

• <65: observed is >20 years, relative is >20 years

• >65: observed is 12.9 years, relative is >20 years,

Table 1 Summary

Chart T1 is a summary synopsis of Table 1 showing comparative indices of mortality, observed mean mortality rate q, with corresponding MR, EDR, and survival SR utilizing 5 risk categories by age and disease duration in years in 2 cohort investigative time-periods, with select follow-up short-and-long-term durational intervals, 0-1, 2-5, 15-20 years abstracted from Table 1. Overall survival and mortality indices are included for the entire 1975-2019 investigational-period, all ages combined.

Table 2: Breast Cancer, Local Stage, Grades 3&4 Combined.

Female breast cancer mortality & survival experience is shown for 67,291 patients, local stage, grades 3&4 (poorly or undifferentiated) in 3 age groups <65, 65 up & all ages combined in both cohorts, 1975-1997 and 1998-2019. At durational intervals from diagnosis of 0-5 to 15-20 years, MR ranged from 119% to 175% with corresponding EDRs of 14.4/1000 at interval 0-5, to 6.5/1000 at 15-20 years and relative cumulative survival SR 82.6% in the last interval. Maximum mortality occurred at the 2-5 year disease-duration interval with observed mean q 0.0379 and corresponding MR 189%, EDR 17.9/1000. Maximum SR 99.7% occurred in the 1^st interval then declined to 82.6% at 20 years. Mortality rates, q & q’, increased with advancing age in both investigative cohorts.

1975-1997, 0-1 year:

• <age 65, observed mean q 0.0093; expected mean q’ 0.0048,

• >age 65, observed mean q 0.0471; expected mean q’ 0.0453.

1998-2019, 0-1 year:

• <age 65, observed mean q 0.0078; expected mean q’ 0.0043,

• >age 65, observed mean q 0.0409; expected mean q’ 0.0402.

Median survival times:

• 1975-1997: observed 16.6 years, relative >20 years,

• 1998-2019: observed >20 years, relative >20 years,

• 1975-2019: observed 19.0 years, relative >20 years.

(Fig 6)

1975-1997, All Ages Combined; there were 24,069 female entrants in this initial cohort. Relative cumulative survival rates (SR=100P/P’) at 2-5 years disease duration is 90.4%. Excess mortality is much higher in localized breast cancer, grades 3&4 (poorly or undifferentiated) as shown in Table 2. For all ages combined, MR rose from 116% in the first year to a peak of 202% at duration 2-5 years and then declined to 119% at 15-20 years. EDR shows a similar pattern at 3.4/1000 in the 1^st interval, rising to a peak of 24.2/1000 at 2-5 years, declining to 6.8/1000 at 15-20 years. Excess mortality, present in the first year in both age groups, was relatively modest in patients age 65 up with all MR values under 135%. After 5-years disease-duration, MR and EDR decreased moderately. However, cumulative relative survival (SR) after 15 years dropped to 78.3% from a high of 90.7% in grade 1&2 disease. Mean annual observed mortality rate q, at 2-5 years was 0.0479; corresponding MR 202%, EDR 24.2/1000, and SR 90.4%. Median observed and relative survival was 16.6 and >20 years, respectively.

1975-1997, Ages <65; duration 2-5 years, MR shows the usual pattern of decrease with advancing age, from 558% at age <65 to 131% at ages 65 up and 5-year SR 89.0%. Maximum observed mean mortality rate q per 1000 at 2-5 years is 0.0345 but declined to 0.0246 in the 15-20 years interval with terminal MR 129%, EDR 5.5/100 and SR 78.3%. Median observed and relative survival indices were >20 years.

1975-1997, Ages 65 up; In the oldest age group in the 2^nd durational interval, the maximum MR is 134%, EDR 16.3/1000 and SR 98.1% based on 575 interval deaths, but this occurs much earlier at 1-2 years from diagnosis; 5-year SR is 93.0%. Observed mean mortality rate q at 2-5 years is 0.0712 indicating 52% larger observed mean q with advancing age, but then increased to 0.1353 at 15-20 years disease-duration based on 1296 deaths (d) and circumscribed exposure (E) 9,581.0 person-years (q = 1296/9,581.0). Median observed & relative survival are 9.2 and >20 years, respectively.

1998-2019, All Ages Combined; There were 43,222 female entrants with local stage breast cancer, grades 3&4 combined in the 1998-2019 cohort. Prognosis is less favorable. An MR of 181% occurred at 1-2 years disease-duration with corresponding EDR 13.6/1000 and SR 98.4%. Significant mortality persists to 15-20 years based on 3855 deaths, maximum observed mean mortality rate q 0.0372, MR 116%, EDR 5.1/1000; SR 86.2%. Both observed & relative median survival were >20 years. Mean annual observed mortality rate q at 2-5 years, 0.0324 indicating 32% mortality rate improvement in the 2^nd investigational time-period; corresponding MR 180%, EDR 14.4/1000, and SR 94.1%.

1998-2019, Ages <65; At duration 2-5 years, MR shows the usual pattern of decrease with advancing age, from 370% at age <65 to 128% at ages 65 up and 5-year SR 94.0%. Maximum observed mean mortality rate q at 2-5 years 0.0204, then declined to 0.0198 at 15-20 years disease with corresponding MR 119%, EDR 3.2/1000 and SR 87.9. Median observed & relative survival >20 years.

1998-2019, Ages 65 up; In the oldest age group, the maximum MR is 136%, EDR 15.5/1000 in the 1-2 years disease-duration. Relative cumulative survival (SR) at 5 years is 94.4% based on 2087 interval deaths. After 15 years, MR and EDR remain elevated at 112% & 14.1/1000 respectively, and SR declined to 77.7%. Mean annual observed mortality rate q per 1000 at 2-5 years is 0.0602 indicating 56% increase in mortality rate q with advancing age and then continues to increase to 0.1311 at 15-20 years disease duration. Median observed & relative survival 10.5 and >20 years, respectively.

1975-2019, All Ages Combined: In a total of 67,291 women with breast cancer, local stage, grades 3&4 (poorly or undifferentiated), all ages combined, excess mortality (EDR) increases to 2-5 years then diminishes to 15-20 years. At durational intervals from diagnosis MR ranged from 116% at 0-1 years, to a maximum of 190% at 1-2 years, decreasing to 119% at 15-20 years disease duration. Overall, in local stage, grade 3&4 breast cancer, increasing disease-duration has a variable impact on mortality. In the last interval, SR associated with histologic grades 3&4 fell to 82.6% from 94.5% associated with grades 1&2 disease. Median observed and relative survival were 19.0 and >20 years, respectively.

Table 2 Summary

Chart T2 is a summary synopsis of Table 2 showing comparative indices of mortality, observed mean mortality rate q, with corresponding MR, EDR, and survival SR utilizing 5 risk categories by age and disease-duration in years, in 2 cohort investigative time-periods, with select follow-up short-and-long-term disease-durational intervals, 0-1, 2-5, 15-20 years abstracted from Table 2. Overall survival and mortality indices are included for the entire 1975-2019 investigational-period, all ages combined.

Table 3: Breast Carcinoma, Regional Stage, Grades 1&2

Female breast cancer mortality & survival experience is shown for a total of 54,921 patients, regional stage, grades 1&2 (well or moderately differentiated) in 3 age groups <65, 65 up & all ages combined in both cohorts, 1975-1997 and 1998-2019. Significant excess mortality existed in regional stage cancer for all durational intervals from diagnosis. Maximum observed mean q for the 1975-2019 cohort 0.0466 occurred at the terminal 15-20 years interval with corresponding MR 158%, EDR 17.1/1000 and SR 68.1%. Maximum observed mean q for the 1975-1997 cohort 0.0582 occurred in the 2-5 year interval with corresponding MR 278%, EDR 37.2/1000 and SR 85.9%. Maximum observed mean q for the 1998-2019 cohort 0.0418 did not occur until the terminal 15-20 year interval with corresponding MR 154%, EDR 14.7/1000 and SR 73.4%. Mortality rates, q & q’, increased with advancing age in both investigative cohorts.

1975-1997, 0-1 years:

• <age 65, observed mean q, 0.0121; expected mean q’, 0.0047,

• >age 65, observed mean q, 0.0502; expected mean q’, 0.0408.

1998-2019, 0-1 years:

• <age 65, observed mean q, 0.0061; expected mean q’, 0.0040,

• >age 65, observed mean q, 0.0384; expected mean q’, 0.0365.

Median survival times-all ages:

• 1975-1997: observed 12.9 years, relative >20 years,

• 1998-2019: observed 19.0 years, relative >20 years,

• 1975-2019; observed 16.8 years, relative >29 years.

(Fig 7)

1975-1997, All Ages Combined: There were 15,149 patients in this 1^st cohort, all ages combined.

Relative cumulative survival rates (SR=100P/P’) at 2-5 years disease-duration is 85.9%. Excess mortality is much higher in regional stage breast cancer, MR rose from 144% in the 1^st year to a peak of 278% at duration 2-5 years and then declined to 159% at 15-20 years. EDR shows a similar pattern at 8.2/1000 in the 1^st interval rising to a peak of 37.2/1000 at 2-5 years, then declining to 18.7/1000 at 15-20 years. Five-year cumulative relative survival SR was 85.9% at 2-5 years, then declined to 58.5% at 15-20 years.

1975-1997, Ages <65 years: There were 9171 female patient entrants age <65 years in this investigative cohort with 111 deaths at 0-1 years. Initial observed mean q was 0.0121 with corresponding MR 258%, EDR 7.4/1000, and SR 99.3%. Observed mean q rose to 0.0361 at 15-20 years with corresponding MR 220%, EDR 19.7/1000, and SR 59.8%. Maximum MR in regional stage, grades 1&2, was as high as 711% at 2-5 years. By contrast, maximum MR in regional stage, grades 3&4, was much higher at 2,022% at 1-2 years. Lowest MR and EDR values were in the first year after diagnosis in both age cohorts. MR at 2-5 years shows the usual pattern of decrease with advancing age, from 711% at age <65 to 185% at ages 65 up. Observed mean q at 2-5 years increased from 0.0416 at ages <65 to 0.0869 at ages >65 years, a 52% increase with advancing age and corresponding MR 185%, EDR 39.8/1000, SR 85.0%. Median observed & relative survival indices, ages <65, 19.2 and >20 years, respectively.

1975-1997, Ages >65 years: There were 5978 female patient entrants age >65 years in this investigative cohort with 300 deaths at 0-1 years. Observed mean q in the initial interval was 0.0502 with corresponding MR 123%, EDR 9.4/1000, and SR 99.0%. Observed mean q then rose to 0.1040 at 15-20 years with corresponding MR 138%, EDR 28.5/1000, and SR 60.4%. Maximum MR 185% & EDR 39.8/1000 occurred at 2-5 years. MR at 2-5 years shows the usual pattern of decrease with advancing age, from 711% at age <65 to 185% at ages 65 up. Median observed & relative survival indices, ages >65, 8.1 to >20 years respectively.

1998-2019, All Ages Combined: There were 39,772 female entrants with regional stage breast cancer, grades 1&2 combined in the 1998-2019 investigative cohort. Prognosis is improved in this latter cohort with initial interval observed mean q 0.0168 with corresponding MR 114%, EDR 2.0/1000 and SR 99.8%. Maximum excess mortality occurred at the 5-10 years disease-duration interval with MR 199%, EDR 18.6/1000, and SR 85.9%. Median observed survival 19.0 years, relative survival >20 years.

1998-2019, Ages <65 years: There were 26,571 female patient entrants age <65 years in this investigative cohort with 161 deaths at 0-1 years. Initial observed mean q was 0.0061 with corresponding MR 152%, EDR 2.1/1000, and SR 99.8%. Observed mean q rose to 0.0266 at 15-20 years with corresponding MR 181%, EDR 11.9/1000, and SR 76.9%. Maximum MR 367% occurred at 2-5 years declining to 181% in the last interval. MR at 2-5 years shows the usual pattern of decrease with advancing age, from 367% at age <65 to 147% at ages 65 up. Median observed & relative survival indices were both >20 years respectively.

1998-2019, Ages >65 years: There were 13,201 female patient entrants age >65 years in this investigative cohort with 506 deaths at 0-1 years. Initial observed mean q was 0.0384 with corresponding MR 105%, EDR 1.9/1000, and SR 99.8%. Observed mean q rose to 0.1239 at 15-20 years with corresponding MR 131%, EDR 29.5/1000, and SR 59.6%. Maximum MR 147% occurred at 2-5 years based on 2,065 interval deaths, then declining to 131% at 15-20 years. Observed mean mortality rate q at 2-5 years, ages <65, increased from 0.0187 to 0.0618 at ages >65 years indicating a 70% increase in observed mean q with advancing age; corresponding MR 147%, EDR 19.7/1000 and SR 92.8%. Median observed & relative survival 10.2 and >20 years respectively.

1975-2019, Regional Stage, Grades 1&2, All Ages Combined; In a total of 54,921 women patients with regional stage, grades 1&2 breast cancer, with 1,078 deaths in the 1^st interval; overall prognosis is diminished. Initial observed mean q 0.0197 at 0-1 year disease duration rose to 0.0466 at 15-20 years, (58% observed mean q increase with disease-duration to 20 years). MR ranged from a maximum of 221% at 2-5 years to 158% at 15-20 years with corresponding terminal EDR 17.1/1000 and SR 68.1% based on 2,475 deaths. Observed median and relative survival indices are 16.8 and >20 years, respectively.

Table 3 Summary

Chart T3 is a summary synopsis of Table 3 showing comparative indices of mortality, observed mean mortality rate q, with corresponding MR, EDR, and survival SR utilizing 5 risk categories by age and disease-duration in years in 2 cohort investigative time periods, with select follow-up short- and long-term disease-durational intervals, 0-1, 2-5, 15-20 years abstracted from Table 3. Overall survival and mortality indices are included for the entire 1975-2019 investigational-period, all ages combined.

Table 4: Breast Cancer, Regional Stage, Grades 3&4

Female breast cancer mortality & survival experience is shown for a total of 50,932 patients, regional stage, grades 3&4 (well or moderately differentiated) in 3 age groups <65, 65 up & all ages combined in both cohorts, 1975-1997 and 1998-2019. Very high mortality was present in all Table 4 disease-duration intervals from diagnosis. Overall, 1975-2019, regional stage, grades 3&4 combined, maximum mortality existed in the 1-2 year disease-duration interval, observed mean q 0.0837 (83% higher than expected mean q of 0.0144) with corresponding MR 581% & EDR 69.3/1000; observed mean q then declining to 0.0385 in the terminal follow-up interval. Maximum relative cumulative survival, SR 97.1% occurred in the 1^st duration, then declined to 54.2% at 20 years. Mortality rates and excess mortality increased precipitously with advancing age:

1975-1997, 0-1 year:

• <age 65, observed mean q, 0.0326; EDR 28.2/1000,

• >age 65, observed mean q, 0.0893; EDR 48.6/1000.

1998-2019, 0-1 year:

• <age 65, observed mean q 0.0194; EDR 15.7/1000,

• >age 65, observed mean q 0.0847; EDR 45.7/1000.

Median survival times-all ages:

• 1975-2019: observed 12.2 years, relative >20 years,

• 1975-1997: observed 8.2 years, relative >20 years,

• 1998-2019: observed >20 years, relative >20 years.

(Fig 8)

1975-1997, All Ages Combined: There were 20,502 female patient entrants with initial observed mean q 0.0514. The excess death rate, EDR, was as high as 88.7/1000 in the 2^nd interval with MR 631%, observed mean q 0.1054. SR was 87.7%, then declined to 44.8% at 15-20 years. Excess mortality is much higher in regional stage, grades 3&4 breast cancer, MR rose from 313% in the 1^st year to a peak of 631% at 1-2 years interval duration and then declined to 162% at 15-20 years. Maximum observed mean mortality rate q 0.1054 in 1975-1997 occurred in the 1-2 year disease-duration interval, then declined to 0.0421 at 15-20 years. MR at 1-2 years shows the usual pattern of decrease with advancing age, from 1394% at age <65 to 287% at ages 65 up.

1975-1997, Ages <65 years: There were 13,727 female patient entrants age <65 years in this investigative cohort with 448 deaths at 0-1 years. Initial observed mean mortality rate q was 0.0326 with corresponding MR 742%, EDR 28.2/1000, SR 97.2%; observed mean q them declined to 0.0286 at 15-20 years duration. Maximum MR 2,022% occurred in the 1-2 year durational interval then declined to 197% at 15-20 years, corresponding EDR 14.1/1000 and SR 46.9%.

1975-1997, Ages >65 years: There were 6775 female patient entrants age >65 years in this investigative cohort with 605 deaths at 0-1 years. Observed mean q in the initial interval was 0.0893 with corresponding MR 219%, EDR 48.6/1000, and SR 94.9%. Observed mean q then rose to 0.1278 at 1-2 years with corresponding MR 300%, EDR 85.2/1000, and SR 86.5%. Maximum MR 300% & EDR 85.2/1000 occurred at 1-2 years. MR at 1-2 years shows the usual pattern of decrease with advancing age, from 2,022% at age <65 to 300% at ages 65 up. Median observed & relative survival indices, ages >65, 5.4 to 10.6 years, respectively.

1998-2019, All Ages Combined: There were 30,430 female patient entrants with regional stage breast cancer, grades 3&4 combined in the 1998-2019 investigative cohort. Initial observed mean mortality rate q 0.0365 indicating a 29% improvement in the 1998-2019 cohort investigative period. Maximum observed mean mortality rate, q, 0.0691 occurred in the 1-2 year disease-duration interval with corresponding MR 540%, EDR 56.3/1000 and SR 92%, then declining in the terminal interval to 0.0315 with corresponding MR 147%, EDR 10.1/1000 and SR 62.9%.

1998-2019, Ages <65 years: There were 22,447 female patient entrants <65 years in this cohort with 434 deaths in the 1^st interval, observed mean mortality rate q 0.0194 with corresponding MR 524%, EDR 15.7/1000 and SR 98.4%. Maximum observed mean q 0.0543 occurred in the 1-2 year interval with corresponding MR 1394%, EDR 50.4/1000 and SR 93.4, then declining to 0.0214 in the 15-20 year interval with corresponding MR 169%, EDR 8.7/1000 and SR 66.2%.

1998-2019, Ages >65 years: There were 7983 female patient entrants >age 65 years in this cohort with 675 deaths in the 1^st interval, observed mean mortality rate, q, 0.0847, MR 217%, EDR 45,7/1000, and SR 95.2%. Maximum observed mean mortality rate q 0.1137 occurred in the 1-2 year interval with corresponding MR 287%, EDR 74.1/1000, and SR 87.9%. Observed mean mortality rate, q, declined to 0.0795 at 5-10 years. but rose to 0.1109 at 15-20 year duration with corresponding MR 122%, EDR 20.4/1000, and SR 49.8. Median observed and relative survival 7.0 and 19.9 years, respectively.

1975-2019, Regional Stage, Grades 3&4,All Ages Combined: In a total of 50,932 women with breast cancer, overall mortality is quite high. At durational intervals from initial diagnosis MR ranged from 297% to a maximum of 581% at 1-2 years then declined to 158% at 15-20 years; EDRs of 28.2 at 0-1 years, 69.3 at 1-2 years, and 14,1/1000 at 15-20 years based on 27,224 total deaths for the entire time-period; In the 2-5 year duration-interval, SR noticeably dropped from 92.0% in grade 1&2 disease to 75.0% in grade 3&4 disease. Terminal duration SR dropped to 54.2%. Median observed & relative survival indices 12.2 and >20 years, respectively.

Table 4 Summary

Chart T4 is a summary synopsis of Table 4 showing comparative indices of mortality, observed mean mortality rate q, with corresponding MR, EDR, and survival SR utilizing 5 risk categories by age and disease duration in years, in 2 cohort investigative time-periods, with select follow-up short-and-long-term disease-durational intervals, 0-1, 2-5, 15-20 years abstracted from Table 4. Overall survival and mortality indices are included for the entire 1975-2019 investigational period, all ages combined.

Table 5: Breast Carcinoma, Distant Stage: All Ages & Grades Combined

Female breast cancer mortality & survival experience for distant stage, all grades combined is shown in 3 age groups <65, 65 up, all ages combined in both cohorts, 1975-1997 and 1998-2019. Observed mean mortality rates, q, were exceedingly high in both cohorts with maximum q in the 1975-1997 cohort exceeding 0.3000 in the 1-2 and 2-5 year durational intervals. In the 1^st interval of the early cohort, observed mortality rate q was 0.3359 with corresponding MR 1,631%, EDR 315.3/1000, and SR 67.8%. In the 1998-2018 cohort, observed mortality rate q in the 1^st interval had diminished to 0.2266 with corresponding MR 1,373%, EDR 210.1/1000, and SR 78.6%. Excess mortality was extremely high in both cohorts with no distinction made by grade, although <age 65. at 15-20 years duration, MR had decreased to 301% in the early cohort and 307% in the later cohort. Unlike the results for local and regional stages, MR was generally highest in the first 2 years after diagnosis and then decreased with greater disease duration. EDR values were generally higher in the >65 age group, in which initial EDRs were 372.8/1000 in 1975-1997, and 296.8/1000 in 1998-2019. Initial MRs in those <65 exceeded 5000% in the 1975-1997 cohort, a 5-fold increase in annual mortality from those ages >65 years and reached almost 4000% in the 1998-2019 cohort. SR for 1975-1997 female entrants was a dismal 8.6% at 15-20 years with only slight improvement to 16.0% in 1998-2019. Only 837 of the original 17,404 entrants remained alive at the 15-20 year follow-up interval. Rapid attrition of survivors is evident in these results for the distant stage of female breast cancer. Median observed and relative survival was 2.5 and 2.7 years respectively. (Fig 9)

1975-1997, Distant Stage: There were 5,708 female patients in this early cohort, all ages & grades combined. More than 79% (4,526) of the patients had died in the first 5-years from initial diagnosis, and more than one-third expired in the initial 0-1 year durational interval. At 2-5 years, observed mean mortality rate q was 0.2414 with corresponding MR 1,126%, EDR 220/1000, and relative cumulative survival (SR) 22.9%. Only 414 patients remained alive in the 15-20 years durational interval. Median observed and relative survival was 1.8 & 1.9 years, respectively.

1998-2019, Distant Stage: There were 11,766 in this latter cohort with only minimal improvement in mortality and survival indices indicating token improvements in diagnosis and treatment for distant stage disease; 7822 patients died (66%) in the first 5 years from initial diagnosis and 23% died in the initial 0-1 year durational interval, slight improvement from the early cohort. MR at 2-5 years was 1207%, corresponding EDR 165.5/1000, and SR 37.1%. In the 1998-2019 investigative cohort at 2-5 years duration, observed mean mortality rate q had diminished to 0.1805, a 25% improvement from a q of 0.2414 in the 1975-97 cohort. Median observed and relative survival had improved to 3.0 and 3.2 years, respectively.

Table 5 Summary

Chart T5 is a summary synopsis of Table 5 showing comparative indices of mortality, observed mean mortality rate, q, with corresponding MR, EDR, and survival SR utilizing 5 risk categories by age and disease-duration in years, in 2 cohort investigative time-periods, with select follow-up short-and-long-term disease-durational intervals, 0-1, 2-5, 15-20 years abstracted from Table 5. Overall survival and mortality indices are included for the entire 1975-2019 investigational-period, all ages, grades combined.

Table 6: Cohort, Stage, Grade, Durations 0-5 & 5-10 Years, All Ages Combined

In Table 6, results are juxtaposed, line by line, for the two cohorts by disease-duration, 1975-2019, female sex, all ages combined, by each combination of stage and grade in the preceding tables, with addition of results for the ungraded cases. Data are aggregates for duration 0-5 years in the top part of the table, and for duration 5-10 years in the lower part. The basic observed data, and both d and P indices are shown as well as the derived MR, EDR, and SR (survival ratio). These data provide a compact overall summary of MR, EDR and SR for disease-durations 0-5 and 5-10 years, all ages combined in complementary support of Tables 1-5 that also display data for all ages.

Comparison of MR and EDR by cohort for each of the groups shows that in every group excess mortality has decreased from 1975-1997 to 1998-2019. There has been a converse increase in SR. For example, in regional cancer grades 1-2, duration 0-5 years, MR decreased from 241% to 170%, EDR decreased from 28.5 to 11.2 per 1000, and SR increased from 85.9% to 94.3%. Note that overall mortality and survival were normal (the same as in the matched US population), for patients with local breast cancer, grades 1&2, in 1998-2019. Note also the predominance of ungraded cases (61,921), about 50% of the total (labeled G&U), in the 1975-1997 cohort, but reduced to about 38% in the 1998-2019 cohort. Mortality in the ungraded cases is intermediate between the rates for grades 1&2 and grades 3&4, as would be expected. The same is true for the total cases, regardless of grading (G&U). The aggregate numbers for E, d and d’ are very large for these 5-year duration intervals, all ages combined, despite subdivision by the stage/grade/cohort combinations. The 95% confidence limits are correspondingly narrow.

Table 7: Sex & Cohort, Mortality and Survival

Male & female breast cancer mortality and survival data, all ages, stages and grades combined, are shown in Table 7 in their combined entirety 1975-2019, and separately for each investigative cohort. Available for survival analysis in the entire 1975-2019 SEER database, there was a total of 538,959 patient entrants, male 3608 (0.7%) and female entrants 535,351 (99.3%). Total numbers of patient-entrants by sex are displayed in Table 7. Significant excess mortality existed in all durational intervals for males and females and both sexes combined in the 1975-2019 investigative period but was somewhat reduced in both sexes in the 1998-2019 sub-cohort. Overall, in the entire 1975-2019 investigative time-period, observed mean mortality rate, q, at 2-5 years disease duration was 0.0715 and 0.0455 in males and females respectively indicating a 34% difference between the sexes; median observed survival time in males and females was 9.5 and 16.1 years respectively indicating a 41% difference in these parameters. (Figs 10, 11)

Table 7 Summary

Chart T7 is a synopsis of Table 7 showing comparative indices of mortality, observed mean mortality rate q, with corresponding MR, EDR, and survival SR utilizing 3 disease-duration risk categories for males, females and both sexes combined, with select short-and-long-term follow-up disease-durational intervals, 0-1, 2-5, 15-20 years, and all ages, stages and grades combined, for the entire 1975-2019 time-period, abstracted from Table 7.

Gender Mortality Risk: Of the 526,351 female breast cancer patients in Table 7, 1975-2019, 226,000 (42.1%) died, and of the total of 3608 male breast cancer patients in the same time-period, 2054 died (57.0%). There were significant differences in the risk of mortality between genders in the overall population. Arias and Xu, authors of the United States Life Tables, 2019 published 2022,²¹  indicated that the difference in life expectancy between the sexes was 5.1 years in 2019, increasing 0.1 year from 2018. The increasing gap during these years is attributed to increases in male mortality due to ischemic heart disease and lung cancer, both of which increased largely as the result of men’s early and widespread adoption of cigarette smoking.

Table 8: Race/Ethnicity, 1975-2019, Mortality and Survival

Of the total number of 535,351 female patient-entrants by race, all ages, stages and grades combined, available for survival analysis in the SEER database, 1975-2019, White females included 445,519 patients (83.2%), Black females 38,168 (7.1%), Other (American Indian/Alaska Native, Asian/Pacific Islander) 49,713 (9.3%) and Unknown 1951 (0.4%). Overall observed mean mortality rates q by race are graphed and presented in Figures 12-14. Differences in observed mean mortality rate q with corresponding MR, EDR, and SR values in three disease-duration intervals, (0-1), (2-5) and (15-20) years are summarized and compared by race and risk category in Chart T8. See Figures 12-14 for graphic presentations of racial observed & expected mortality rates by disease-duration.

Race/Ethnicity Mortality Risk: In Table 8, of the 535,351 female breast cancer patients, 1975-2019, 445,519 were White of which 195,269 (43.8%) died. Of the total of 38,168 Black female patient-entrants, 16,399 (43.0%) died. Of the total of 49,713 Other female patients, 14,083 (28.3%) died.

White, 1975-2019, (2-5) years: There were 382,618 White female entrants, 1,043,603.5 patient-years exposure (E), and 47,737 observed interval deaths (13% of White interval-entrants) with corresponding observed mean mortality rate q 0.0457, MR 206%, EDR 23.5/1000 and SR 88.8%. In the short-term, maximum mean mortality rate q in Whites occurred in the 2-5 year disease-duration interval, 0.0457, and in the long-term, 15-20 years, 0.0462.

Black, 1975-2019, (2-5) years: There were 30,068 Black female entrants, 78,615.5 patient-years exposure (E), and 4,806 observed interval deaths (16% of Black entrants in the 2-5 year interval) with corresponding observed mean mortality rate q 0.0611, MR 340%, EDR 43.1/1000 and SR 78.9. However, the maximum observed mean mortality rate in Blacks occurred in the 1-2 year disease-duration interval, 0.0728, with corresponding MR, EDR and SR of 441%, 56.3/1000, and 90.2% respectively. (In the 2^nd, 1-2 year interval, observed mean annual mortality rate q in Blacks was 38% higher than in Whites and 54% above that in Other/ethnicities).

Other (American Indian/Alaska Native, Asian/Pacific Islander), 1975-2019, (2-5) years: There were 41,466 Other female entrants in the 2-5 year interval, 110,740.0 patient-years exposure (E), and 3592 observed interval deaths (9% of Other interval entrants) with corresponding observed mean mortality rate-q 0.0324, MR 346%, EDR 23.1/1000, and SR 89.2%.

Mortality and survival outcomes for Other were between that of Whites and Blacks as might be expected. Not shown here, in separate mortality and survival analyses by investigative time-period, significant improvement in observed mean mortality rates, q, in the 1998-2019 cohort by race and ethnicity occurred. The ethnicities comprising ‘Other’ are a heterogenous group and require further individual study. However, these racial disparities are further discussed in the Commentary section of this study.

Table 8 Summary

Chart T8 is a summary synopsis of Table 8 showing comparative indices of mortality including, observed mean mortality rate q with corresponding MR, EDR, and survival SR by race/ethnicity for 1975-2019. Three risk categories abstracted from Table 8 with select short-and-long-term follow-up disease-durational intervals 0-1, 2-5, 15-20 years were used to compare mortality & survival outcomes for: All races, White, Black, Other (American Indian/Alaska Native, Asian/Pacific Islander), and Unknown, all ages, stages and grades combined.

Table 9: Selected Oncophenotypes, M&S, All Ages-Stages-Grades Combined

Chart T9 is a synopsis of Table 9 showing comparative indices of mortality and survival including: Table references, total numbers of entrants with intervals of 0-5 year disease duration, aggregate average annual values for MR, EDR, SR, and for selected ICD-O-3 histologic onchophenotype breast cancers by sex and race in the 1998-2019 or other time-periods as noted, all ages, stages & grades combined:

• Carcinoma in-situ; Seer historic Stage A indices, (abstracted from Table 9),

• Carcinoma in-situ; SEER adjusted AJCC indices, (6^th edition), abstracted from Table 11,

• 6 results for ICD-O-3 histologic oncophenotypes,

• 2 results by sex,

• 3 results by race.

In Table 9, Invasive intraductal carcinoma is by far the predominant ICD-O-3 histologic onchophenotype (207,902 cases), lobular carcinoma (25,141 cases), and all others much less. Carcinoma in-situ, abstracted from SEER historic stage A data, spreadsheet Table 9, with 61,716 cases, displayed no excess mortality with MR in every interval ranging from 41 to 95% and corresponding (negative) EDRs -6.8 to -1.2/1000 person-years. SR was slightly above 100% in every duration. Curiously, carcinoma in-situ abstracted from adjusted AJCC 6^th edition data with only 377 cases, spreadsheet Table 11, did display very slight excess mortality within the 2-5, 5-10, and 10-15 year intervals.

At duration 0-5 years in the previous Table 8, both MR and EDR are higher in Black patients than Whites by a considerable margin but the differences in observed mortality rate q are smaller at duration 5-10 years. In the summary synopsis Chart T9, 1975-2019, SR in Black patients at 5 years disease-duration is 78.9% and in White patients 88.8%. Black female breast cancer patients observed interval mortality rate q for the entire duration of the 0-5-year interval is 0.0632, much higher than that for White female breast cancer patients 0.0441, a 1.4 fold elevation in risk. Observed mean mortality q is variable with advancing durations. Differences in mortality and survival have persisted between both races through 2019 in the United States. Mean age in blacks, 57.9 in 1998-2019, is much lower than in whites, 62.6 years.

Table 10: Breast – SEER Adjusted AJCC 6^th Ed. (1988-2015) Stages I-IV; All Ages and Grades Combined

In the 1998-2019 cohort time-frame, Table 10 shows results for female breast cancer classified by the SEER adjusted AJCC 6^th Edition extent-of-disease system of stages I-IV, defined by corresponding combinations of TNM (Tumor, Node, Metastasis) iterations. See page 2 for the AJCC outline-plan of 4 stages and multiple TNM group variations for breast cancer. Stages IIIC and IV are the only stages that include cases with distant metastasis (M1). All other cases are divided into 3 stages with Roman numeral designations instead of the 2 stages (local and regional) of the historic staging system.

• Stage I includes subcategories IA & IB cases with tumor size not over 2 cm (T0 or T1) and tumor-free nodes-N0) or nodes with tumor N1.

• Stage II includes subcategories IIA & IIB cases with tumor 2-5 cm (T2), and N0 or N1 nodes.

• Stage III includes subcategories IIIA, IIIB, IIIC cases with tumor size greater than 5 cm (T3), and any tumor with direct extension to tissues outside the breast (T4); The 3 Stage III sub-categories contain a total of 10 diverse TNM groupings.

• Stage IV is exclusive to cases with distant metastasis (M1) only.

In Tables 10 & 11, each AJCC staging band incorporates all of the TNM grouping iterations pertinent to the stage. As it does with the historic staging system, excess mortality in the AJCC staging scheme increases progressively from stage I to stage IV. For interval duration 2-5 years the EDR increases from -1.6/1000 in stage I, to 15.6 in stage II, to 58.2 in stage III, to 210.5 in stage IV. In the lower part of Table 10, comparison of Adjusted AJCC Stages I-IV, all combined, are compared with SEER Historic Stage A, local, reginal, distant (LRD), also combined, shows remarkable similarity in excess mortality (EDR) and survival (SR) for interval duration 2-5 years; AJCC EDR 15.1/1000 & SR 92.5%, and SEER 15.7/1000 & 92.1%, respectively. In Table 10, AJCC staging, as in the SEER historic Stage Table 1, the same curious pattern existed within one year of diagnosis, all ages combined, of observed q 0.0082 less than expected q’ 0.0163 with MR 50%, and, at 2-5 years duration, MR of only 92%.

Table 11: Breast – SEER Adjusted AJCC 6^th Ed. (1988-2015); Stages 0 (Tis), I, II; Grades 1&2 and 3&4

In the 1998-2019 cohort time-frame, Table 11 shows the subdivision of stages I and II by our usual grade combinations. These reflect the same differentiation seen previously for grading in local breast cancer. In stage I, for example, EDR at duration 2-5 years is -3.6 in grades 1&2 and in grades 3&4 it is 5.8 per 1000. The weighted average is -1.7 all ages and grades 1-4 combined. Grading is therefore highly successful in distinguishing EDR values in a given stage/age/duration combination. The AJCC staging system itself reflects the importance of tumor size as a factor in assessing prognosis.

Tables, Charts and Figures: It is important to remember that most of the calculations were performed to develop complete breast cancer mortality tables of observed and derived data in progressive intervals of time to determine parameters of mortality and survival probabilities: Mortality rates (q), mortality ratios (MR), excess death rates (EDR), and relative survival rates (SR), and their morbid relationship to enumerated breast cancer risk factors. Eleven mortality tables display and mathematically support the medical-actuarial short- and long-term breast cancer outcomes analyzed in this study. Charts are organized medical-actuarial summaries, abstracted from the 11 mortality tables for ease of data comparison and support of this analytic narrative. Figures, visual diagrams that clarify and support epidemiologic & biostatistical cancer outcomes, underscore the importance of annual mortality rates (q) as the basis for breast cancer mortality tables.

Observed Mean Mortality Rates (q): Observed mean mortality rates (q) are weighted by the frequency of occurrence of death (d) in a defined population at risk of death (exposure E in patient-years) during a specified interval (q = d/E) and are emphasized in this study as the preferred comparative measure of progress and prognosis in breast cancer. The quotient of the d/E ratio is understood by convention as the observed mortality rate q and is typically expressed as a decimal in units of deaths per 1000 individuals per year. As the quotient of the d/E ratio increases, the mortality rate q, by definition, increases. The questions then arise:

1. In a 20-year follow-up study why do breasvt cancer mortality rates q frequently decline in the short term from diagnosis (intervals of up to 5- or 10-years disease duration) and then resume on an upward spiral thereafter as seen inFigure 4 and other graphs in this study?

2. A corollary question: Why is there so much greater mortality associated with older-age, 65 up, vs younger-age onset as seen in the 15-20 year durations ofTables 1-5 and 7-11?

At least 2 plausible explanations may be at work here: (1), It may be that in the long-term and current absence of non-cytotoxic onchophenotype DNA-specific targeted therapies, breast cancer patients, despite alternative, adjunctive treatments to slow the disease in the short-term, remain exposed to the biologic ravages of continuing unabated or ameliorated DNA/molecular distortions with deadly consequences later; (2). In normal cells, the gene for transforming growth factor β, (TGF-β), acting through its signaling pathway, stops the cell cycle at the G1 stage to stop proliferation, induce differentiation, or promote apoptosis. In many cancer cells, parts of the TGF-β signaling pathway are mutated, and TGF-β no longer controls the cell. These cancer cells proliferate. In a landmark study published in JAMA 2001,²²  Elad Ziv, MD and associates examined the relationship between a functional polymorphism in the TGF-B1 gene and breast cancer risk among a cohort of elderly white women ages 65 or older with 9.3 years mean follow-up. They concluded that: 1) TGF-B1 gene C/C allele, (the normal allele) is protective unless it mutates, and 2) Mutated TGF-B1 gene polymorphic alleles T/T and T/C appear to express themselves only at longer durations in older age-onset women with breast cancer; thus, the intense and critical need for continuing research.

Table Methodology: Detailed comparative mortality and survival results are given in Tables 1-11 for a variety of combinations of age, sex, race/ethnicity, stage, grade, disease-duration and cohort investigative period. All Tables display breast cancer disease-duration intervals in years of aggregate average annual exposures to the risk E, mean annual observed & expected mortality rates, q & q’, and corresponding cumulative observed and expected survival, P & P’, excess mortality & cumulative relative survival as MR, EDR and SR. The range of these indices is very wide.

Breast Cancer Mortality Risks: Mortality and survival outcomes are succinctly summarized for ease of inspection, mortality comparison and study in Charts 1-5 and 7-9, each Chart numerically corresponding with the appropriate spreadsheet Table for each category of risk. Tables 10 and 11 are the more specific AJCC stage comparisons than the SEER local, regional, distant (LRD) staging system.

Prognosis: Observed mortality rates q increase, but mortality ratios MR, excess death rates EDR and survival SR diminish with advancing age, stages & grades of cancer. Nonetheless, mortality rates may vary somewhat with advancing disease-duration in some analytic groupings. However, prognosis is improved in the current investigative-period, 1998-2019, because of continuing latter-day research progress, earlier diagnosis & more effective treatment. Prognosis is also impacted by secular particularities (Social Determinants of Health-SDOH) in availability of quality, access and affordability of medical care services per selected regional features of its organization, delivery and finance, besides racial/ethnic/cultural attributions as noted in this and other studies of staging and mortality disparities, as well as biologic properties specific to each cancer onchophenotype – an incredibly diverse and complex mix of prognostic characteristics and variables.

Race-Stage-Mortality Disparity: In the United States, Black women with breast cancer present with more advanced stages and have poorer survival rates than Whites and the heterogenous ethnic group of American Indians/Alaska Natives, Asians/Pacific Islanders, (see Table 8 and Summary Chart T8 for study findings).

Li and colleagues²³  using data from 11 population-based tumor registries that participate in the Surveillance, Epidemiology, End Results Program studied the relationship between race and ethnicity and breast cancer stage, treatment and mortality rates in 124,934 women diagnosed as having a first primary breast cancer between January 1, 1992, and December 31, 1998, that included 97,999 non-Hispanic Whites, 10,500 Blacks, 8834 Asians and Pacific Islanders, and 7219 Hispanic Whites. They found that Blacks and other ethnicities had a 20% to 200% greater risk of mortality after breast cancer diagnosis. They concluded that differences in breast cancer stage, treatment and mortality rates are present by race and ethnicity, and breast cancer survival may be improved by targeting factors, particularly socioeconomic and cultural factors, that underlie these differences.

Similarly, Jemal and colleagues²⁴  in a large national sample of non-elderly white and black women with invasive early-stage breast cancer, concluded that differences in insurance accounted for a substantial proportion of the excess risk of death among black women versus white women. They also concluded that “equalizations of access to care in non-elderly black women could substantially reduce ethnic/racial disparities in overall mortality of women diagnosed with breast cancer.”

In a much wider sense as noted above, and well beyond the scope of this study, the CDC (Center for Disease Control and Prevention) has begun to address the problem of health equity and the social determinants of health (SDOH) – non medical factors that influence health outcomes – by adopting the SDH definition from the World Health Organization (WHO) – conditions in which people are born, grow, work, live, and age, and the wider set of forces and systems shaping the conditions of daily life.²⁵ Godspeed to their efforts.

Actuarial Medicine: This medical-actuarial study of breast cancer mortality risks is an evidence-based approach to better understand and cope with a formidable, complex and tragically morbid cultural-biologic predator, without compassion for humanity by age, sex, race, ethnicity or historical era. The AJCC staging system itself reflects the importance of tumor size as a factor in assessing prognosis. However, in the SEER database it can only be used for cases diagnosed and so coded in 1988 or later. Nevertheless, while awaiting onchophenotype DNA-specific targeted therapies, with current diagnostic and therapeutic advances, advent of progressive genome testing, hormone and cyclin dependent kinase 5-6 inhibitors (CDK 5-6 inhibitors), and other immuno, chemo, medical, surgical and radiologic remedies, the best is yet to come! Optimum Etiamnunc Est Advenio.

Thoughts En Passant: Admiration and respect for the fortitude of untold millions of patient-sufferers at the ignoble hands of ‘this accursed pest,’ courageous convalescents within the purview of my own family, friends, patients, and acquaintances, and gratitude for the academic professionalism and expertise of numerous educators, colleagues and mentors, especially Richard B. Singer, M.D. (March 22, 1914 – February 19, 2010), for imparting his wisdom and the special tools of scientific apprehension and technical mechanics necessary for this study – Excelsior!

Breast Cancer Havoc,

Prologue Past, Epilogue Near,

Ramazzini, Ehrlich, St. Ignatius Loyola Withal,

Prophetic Words Eager to Hear,

Magic Bullets Target This Accursed Pest,

For the Greater Glory of God – AMDG,

And Humankind Salvation – IHS,

Godspeed

The Best is Yet to Come,

Excelsior!

Anthony F. Milano, MD, MPH