Abstract
Background
Although breast cancer is often perceived to be indolent in older women, breast cancer outcomes in the oldest patients are variable. We examined breast cancer-specific death by age, stage, and disease subtype in a large, population-based cohort.
Methods
Using Surveillance, Epidemiology, and End Results data, we identified 486,118 women diagnosed with stage I-IV breast cancer during 2000-2012. Using a series of Fine and Gray regression models to account for competing risk, we examined the risk for breast cancer-specific death by age and stage (I-IV) for sub-cohorts with hormone receptor (HR)-positive, HR-negative, human epidermal growth factor receptor 2 (HER2)-positive, and triple negative disease, adjusting for demographic and clinical variables.
Results
Overall, 18% of women were 65-74, 13% were 75-84, and 4% were ≥85. Regardless of stage within HR-positive and HR-negative cohorts, patients aged ≥75 (vs. 55-64) experienced a higher adjusted hazard of breast cancer-specific death, which was particularly evident for those with early-stage, HR-positive disease (hazard ratio for ages 75-84=1.88, 95% Confidence Interval [CI]=1.68-2.09 and hazard ratio for age ≥85=3.59, 95% CI=3.12-4.13 [both for stage I disease]). In HER2-positive and triple negative cohorts, women age ≥70 had consistently higher risk for breast cancer-specific death across stages (vs. ages 51-60), with exception of stage IV triple negative disease.
Conclusions
Older patients experience worse breast cancer outcomes, regardless of subtype and stage. With an increasing number of older patients anticipated to develop breast cancer in the future, addressing disparities for older patients must emerge as a clinical and research priority.
Keywords: breast cancer, older women, age, disparities, survival
INTRODUCTION
Although breast cancer in older women is often perceived to be a more indolent disease than that seen in younger women because of favorable tumor biology,1 early stages of disease at presentation,2 and competing risks of death,3 the outcomes of older patients with breast cancer are highly variable,2, 3 and it is at the extremes of age where patients are at risk for the worst breast cancer outcomes.2, 4–6 Despite the fact that many older patients are more likely to die of non-breast cancer causes when diagnosed with early stages of breast cancer (particularly when coexisting comorbidity is advanced),3, 7, 8 their breast cancer outcomes can also be worse than their younger counterparts. Currently, women age ≥70 represent 31% of all annual breast cancer cases, yet they account for 47% of all breast cancer-specific deaths.9 Further, the improvements noted in breast cancer survival over time have occurred at a slower rate for those age ≥75 compared with those observed in younger patients.6
Past studies have described differences in survival by age overall4 and within the older populations studied within Surveillance, Epidemiology, and End Results (SEER)-Medicare who have early stage disease.2 Although differences in disease biology and subtypes have been explored by age to some degree4, 5,10 current data are limited on age-related differences in breast cancer-specific survival by stage and disease subtype. To better understand potential disparities for the oldest breast cancer patients, we examined the risk for breast cancer-specific deaths for adult women with stage I-IV breast cancer by age, stage, and disease subtype among women diagnosed with breast cancer during a modern era, 2000-2012.
METHODS
Data Source and Study Population
We identified a cohort of patients in the SEER registry with a first diagnosis of stage I-IV breast cancer. The 17 population-based SEER cancer registries cover areas that represent 28% of the U.S. population and uniformly collect information on patient demographics, tumor characteristics, initial treatment, and mortality for all incident cancers.11 Because this study used registry data, it was deemed exempt for review by the Office for Human Research Studies at Dana-Farber Cancer Institute.
We identified 487,983 women who were diagnosed with their first stage I-IV breast cancer during 01/01/2000-12/31/2012 who had cancer histologies likely to be treated by standard guidelines and who were not diagnosed at autopsy/death. Women with bilateral cancers (due to potential misclassification from differing subtypes and stages; n=1,768), unknown age or age <18 (n=18), and invalid coding for survival months (n=79) were excluded, with a final analytic cohort of 486,118 women. Further sub-cohorts were created to examine outcomes in those with stage I, II, III, (de novo) IV disease and in those with varying subtypes as described below.
Outcome Measures
The primary outcome of interest was breast cancer-specific death, with death due to other causes considered a competing event. The interval of survival time was defined as the date of diagnosis until the date of death from any cause or the date of censoring at the last follow-up date available (12/31/2013). We ascertained deaths and causes of death from National Death Index data provided with the SEER files.
Independent Variables
Our independent variables of interest included age, American Joint Committee on Staging (AJCC) 6th edition breast cancer stage (I, II, III, IV), and breast cancer subtype. Age was defined as ≤35, 36-44, 45-54, 55-64, 65-74, 75-84, and ≥85 years for all models except those examining outcomes for those with human epidermal growth factor receptor 2 (HER2)-positive disease and triple negative disease because HER2 status was only available in 2010-2012 and sample sizes were substantially smaller than other cohorts. In HER2-positive and triple negative models, we categorized age as <40, 40-50, 51-60, 61-70, >70 years. Breast cancer subtype was categorized using hormone receptor (HR) status and HER2 status. For HR status, having estrogen receptor (ER)- and/or progesterone receptor (PR)-positive disease was categorized as HR-positive, having ER- and PR-negative disease was categorized as HR-negative, with the remaining cases classified as HR-unknown. HER2 status (available during 2010-2012) was categorized as positive, negative, unknown/borderline. Triple negative disease was defined for those diagnosed 2010-2012 only and required HR-negative and HER2-negative disease. In cohorts where we restricted to specific disease subtypes (e.g. HR-positive, HER2-positive, or triple negative), those with unknown values for HR or HER2 were excluded.
Control Variables
Control variables included race/ethnicity (non-Hispanic white, non-Hispanic black, Hispanic, other/unknown), year of diagnosis, marital status (married, single/separated/divorced/widow, unknown), SEER region (Connecticut, Detroit, Hawaii, Iowa, New Mexico, Seattle, Utah, Atlanta/rural Georgia, California, Kentucky, Louisiana, New Jersey), and tumor grade (low/intermediate, high, unknown).
Statistical Analysis
We first defined cohorts by stage (I-IV) and age group for all 486,118 women. Next, we defined sub-cohorts of women by stage and disease subtype, categorized by stage for those with HR-positive, HR-negative, HER2-positive, and triple negative disease. We then examined the number of deaths and causes of death by age and stage overall and used Kaplan-Meier estimates to describe the overall survival and cumulative incidence of death in each cohort. Inference on breast cancer-specific survival was made using a series of Fine and Gray regression models,12 allowing for a sub-distribution of hazards of death due to breast cancer when considering death due to other causes as a competing event.
We performed sequential Fine and Gray models for each of the 4 stage-specific cohorts, after adjustment by the control variables described above, followed by 16 more models where we repeated analyses for each stage (I-IV) but with the additional categorization by tumor subtype for the following sub-cohorts: (a) HR-positive disease, adjusting for all control variables (except HR status), (b) HR-negative disease, adjusting for all control variables (except HR status), (c) HER2-positive disease (2010-2012 diagnoses), and (d) triple negative disease (2010-2012 diagnoses). For models (c) and (d), we adjusted for all variables except HER2 status.
In the univariate and multivariate competing risk regression models, Wald-type tests were conducted of main effects and interaction terms, using two-sided p<0.05 to conclude statistical significance. Point estimates are reported with 95% confidence intervals (95% CI). All statistical analyses were conducted using SAS version 9.4 (SAS Institute, Cary, NC).
RESULTS
Among the 486,118 women with breast cancer identified, 3% were age ≤35 at diagnosis, 11% were 36-44, 24% were 45-54, 25% were 55-64, 18% were 65-74, 13% were 75-84, and 4% were ≥85. Overall, 47%, 35%, 13%, and 5% presented with stage I-IV disease, respectively. Sample sizes and the distribution of age-groups by cohort are shown in Table 1. Additional patient and disease characteristics by stage and by age-group are summarized in Supplemental Table 1. Briefly, most women were white (65-76% for stage I-IV) and most had HR-positive disease but the prevalence decreased with stage (80% of stage I to 61% of stage IV). Among those diagnosed in 2010-2012, HER2-positive cancers occurred in 11% of stage I, 16% of stage II, 21% of stage III, and 22% of stage IV cancers (Supplemental Table 1).
Table 1.
Cohort Sample Sizes by Age (n, %)a [N=486,118]
| COHORT | Total | AGE (years) | ||||||
|---|---|---|---|---|---|---|---|---|
| All ages | ≤35 | 36-44 | 45-54 | 55-64 | 65-74 | 75-84 | ≥85 | |
| Stage-specific cohorts | ||||||||
| Overall | 486,118 | 13,348(3) | 53,439(11) | 118,690(24) | 122,206(25) | 89,860(18) | 63,953(13) | 20,232(4) |
| Stage I | 226,347 | 3,350(2) | 19,079(8) | 51,367(23) | 60,103(27) | 50,949(23) | 33,236(15) | 8,263(4) |
| Stage II | 169,973 | 6,174(4) | 22,602(13) | 44,694(27) | 40,607(24) | 28,678(17) | 19,847(12) | 7,371(4) |
| Stage III | 65,081 | 3,075(5) | 9,583(15) | 17,596(27) | 15,148(23) | 9,746(15) | 7,024(11) | 2,909(4) |
| Stage IV | 24,717 | 749(3) | 2175(9) | 5,033(20) | 6,348(26) | 4,877(20) | 3,846(16) | 1,689(7) |
| Additional sub-cohorts | ||||||||
| Stage I, HR+b | 181,712 | 2259(1) | 14,443(8) | 40,581(22) | 48,277(27) | 42,079(23) | 27,390(15) | 6,683(4) |
| Stage II, HR+b | 123,225 | 3,639(3) | 15,417(13) | 31,913(26) | 29,661(24) | 21,914(18) | 15,172(12) | 5,509(4) |
| Stage III, HR+b | 43,445 | 1,866(4) | 6,227(14) | 11,727(27) | 10,170(23) | 6,760(16) | 4,797(11) | 1,898(4) |
| Stage IV, HR+b | 15,070 | 454(3) | 1,352(9) | 3,083(20) | 3,997(27) | 3,028(20) | 2,282(15) | 874(6) |
| Stage I, HR−c | 28,211 | 864(3) | 3,221(11) | 7,256(26) | 7940(28) | 5328(19) | 2,946(11) | 656(2) |
| Stage II, HR−c | 35,077 | 2,115(6) | 5,734(16) | 9,980(28) | 8,542(24) | 4,752(14) | 2,957(8) | 997(3) |
| Stage III, HR−c | 16,571 | 1,018(6) | 2,716(16) | 4,666(28) | 3,896(24) | 2,221(13) | 1,472(9) | 582(4) |
| Stage IV, HR−c | 5,251 | 229(4) | 563(11) | 1,267(24) | 1,414(27) | 932(18) | 632(12) | 214(4) |
| Stage I, HER2+d | 6,299 | 175(3) | 690(11) | 1724(27) | 1841(29) | 1206(19) | 520(8) | 143(2) |
| Stage II, HER2+d | 7,022 | 404(6) | 1,016(14) | 1,981(28) | 1,860(26) | 1,008(14) | 545(8) | 208(3) |
| Stage III, HER2+d | 3,240 | 204(6) | 517(16) | 959(30) | 814(25) | 417(13) | 232(7) | 97(3) |
| Stage IV, HER2+d | 1,486 | 80(5) | 170(11) | 369(25) | 443(30) | 243(16) | 125(8) | 56(4) |
| Stage I, triple negativee | 4,718 | 120(3) | 443(9) | 1,040(22) | 1,361(29) | 1,090(23) | 522(11) | 142(3) |
| Stage II, triple negativee | 6,258 | 379(6) | 938(15) | 1,750(28) | 1,604(26) | 889(14) | 497(8) | 201(3) |
| Stage III, triple negativee | 2,314 | 145(6) | 356(15) | 614(27) | 546(24) | 329(14) | 219(9) | 105(5) |
| Stage IV, triple negativee | 790 | 28(4) | 88(11) | 164(21) | 232(29) | 146(18) | 94(12) | 38(5) |
Total percents may not equal 100% because of rounding
Hormone receptor-positive (HR+)=estrogen receptor (ER)-positive or progesterone receptor (PR)-positive disease
HR−(negative)=ER-negative and PR-negative disease
HER2+ (human epidermal growth factor receptor 2-positive)=HER2+ disease (2010-2012 diagnoses)
Triple negative=HR-negative and HER2-negative disease (2010-2012 diagnoses)
A total of 105,058 deaths were observed, of which 57,480 (55%) were related to breast cancer, and 47,578 (45%) were attributed to other causes. Among patients without a death recorded, the median (interquartile range) follow-up time was 6.0 years (3.2-9.5 years). The risk of death over a median of 6 years of follow-up increased with age, occurring in 2,640 (20%) of women ages ≤35; 7,797 (15%) of women ages 36-44; 15,471 (13%) of women ages 45-54; 18,381 (15%) of women ages 55-64; 20,349 (22%) of women ages 65-74; 26,978 (42%) of women ages 75-84; and 13,442 (66%) of women ages ≥85. Causes of death by age and stage are shown in Supplemental Table 2 with a higher proportion of older (vs. younger) patients dying of non-breast cancer causes at each stage.
The overall survival by stage is shown in Figure 1A, and the cumulative incidence of breast cancer-specific deaths and deaths from other causes are shown in Figures 1B and 1C, respectively. The cumulative incidence of deaths due to breast cancer (95% CI) at 5-years was 1.9% (1.8-2.0%) for stage I, 8.0% (7.9-8.2%) for stage II, 26.6% (26.2-27.0%) for stage III, and 70.2% (70.0-70.8%) for stage IV disease, respectively.
Figure 1.
Kaplan-Meier estimates of (A):overall survival and the cumulative incidence of (B):breast cancer specific deaths, and (C):death due to other causes are shown for women with stage I (black), stage II (red), stage III (green) and stage IV (blue) breast cancers diagnosed during 2000-2012 (N=486,118)
Supplemental Figure 1 displays the cumulative incidence of breast cancer-specific death by age-group in each cohort by stage at diagnosis. In all stages, the cumulative incidence was highest for ages ≥85, with 4.1% (3.7-4.6%) for stage I, 13.9% (13.1-14.7%) for stage II, 39.1% (37.4-40.9%) for stage III, and 78.1% (75.7-80.5%) for stage IV, respectively. Cumulative incidences were lower for ages 75-84 and younger age-groups.
In the multivariable Fine and Gray model including all patients (Table 2), the adjusted hazard ratios (95% CI) for breast cancer-specific death by age were the following (all vs. ages 55-64): age ≤35=1.06 (95% CI=1.01-1.11), ages 36-44=0.94 (95% CI 0.91-0.97), ages 45-54=0.91 (95% CI=0.88-0.93), ages 65-74=1.08 (95% CI=1.05-1.11), ages 75-84=1.37 (95% CI=1.33-1.42), and ages ≥85=1.65 (95% CI 1.58-1.73). The relationship between the effects of age and stage was explored as an interaction term in the multivariable Fine and Gray model and reached statistical significance (Wald p<0.001). The modification by stage is shown in Figure 2, with a linear relationship of age seen in stage IV patients, versus a parabolic relationship in stage I patients with worse breast cancer specific outcomes in the oldest and youngest age-groups.
Table 2.
Adjusted hazard ratio for breast cancer-specific survival by cohort (95% Confidence Interval)*
| Cohort | |||||||
|---|---|---|---|---|---|---|---|
| AGE (Years) | |||||||
|
2000-2012 cases (n=486,118) |
≤35 | 36-44 | 45-54 | 55-64 | 65-74 | 75-84 | ≥85 |
| Stage I-IVa | 1.06 (1.01-1.11) | 0.94 (0.91-0.97) | 0.91 (0.88-0.93) | 1.00 | 1.08 (1.05-1.11) | 1.37 (1.33-1.42) | 1.65 (1.58-1.73) |
| Stage Ib | 1.64 (1.39-1.94) | 1.33 (1.20-1.47) | 0.95 (0.87-1.03) | 1.00 | 1.19 (1.09-1.29) | 1.81 (1.67-1.97) | 3.00 (2.67-3.36) |
| Stage IIb | 1.24 (1.15-1.34) | 1.00 (0.94-1.05) | 0.91 (0.87-0.96) | 1.00 | 1.16 (1.10-1.22) | 1.58 (1.50-1.67) | 2.16 (2.01-2.32) |
| Stage IIIb | 1.07 (0.99-1.14) | 0.93 (0.88-0.97) | 0.92 (0.88-0.96) | 1.00 | 1.06 (1.01-1.11) | 1.36 (1.30-1.44) | 1.67 (1.55-1.80) |
| Stage IVb | 0.78 (0.72-0.85) | 0.80 (0.75-0.84) | 0.90 (0.86-0.94) | 1.00 | 1.03 (0.99-1.08) | 1.14 (1.08-1.20) | 1.16 (1.08-1.25) |
| Stage I, HR+c,d | 2.34 (1.88-2.92) | 1.38 (1.20-1.58) | 0.91 (0.81-1.02) | 1.00 | 1.27 (1.15-1.42) | 1.88 (1.68-2.09) | 3.59 (3.12-4.13) |
| Stage II, HR+c,d | 1.56 (1.40-1.74) | 1.05 (0.97-1.13) | 0.91 (0.85-0.97) | 1.00 | 1.20 (1.12-1.29) | 1.65 (1.54-1.78) | 2.29 (2.09-2.52) |
| Stage III, HR+c,d | 1.11 (1.01-1.22) | 0.92 (0.87-0.99) | 0.87 (0.82-0.92) | 1.00 | 1.06 (0.99-1.13) | 1.36 (1.27-1.46) | 1.57 (1.42-1.74) |
| Stage IV, HR+c,d | 0.76 (0.68-0.86) | 0.75 (0.69-0.80) | 0.89 (0.84-0.94) | 1.00 | 1.06 (1.00-1.13) | 1.20 (1.12-1.29) | 1.26 (1.14-1.40) |
| Stage I, HR−c,e | 1.13 (0.85-1.49) | 1.26 (1.07-1.48) | 1.02 (0.89-1.17) | 1.00 | 1.02 (0.88-1.19) | 1.60 (1.36-1.89) | 1.94 (1.46-2.58) |
| Stage II, HR−c,e | 1.02 (0.91-1.16) | 0.96 (0.88-1.05) | 0.93 (0.86-1.00) | 1.00 | 1.16 (1.06-1.27) | 1.59 (1.44-1.75) | 1.93 (1.67-2.23) |
| Stage III, HR−c,e | 1.05 (0.95-1.17) | 0.92 (0.85-0.99) | 0.98 (0.92-1.05) | 1.00 | 1.04 (0.96-1.13) | 1.35 (1.23-1.48) | 1.82 (1.59-2.10) |
| Stage IV, HR−c,e | 0.78 (0.68-0.90) | 0.85 (0.76-0.94) | 0.93 (0.85-1.01) | 1.00 | 1.03 (0.93-1.13) | 1.17 (1.04-1.32) | 1.37 (1.13-1.68) |
| AGE (Years) | |||||||
|
2010-2012 cases (n=126,634) |
– | <40 | 40-50 | 51-60 | 61-70 | >70 | – |
| Stage I, HER2+f,g | – | 1.00 (0.22-4.61) | 0.54 (0.17-1.73) | 1.00 | 2.03 (0.91-4.51) | 3.12 (1.42-6.90) | – |
| Stage II, HER2+f,g | – | 0.82 (0.39-1.74) | 0.71 (0.40-1.28) | 1.00 | 1.71 (1.06-2.76) | 3.81 (2.48-5.85) | – |
| Stage III, HER2+f,g | – | 0.83 (0.52-1.32) | 0.91 (0.63-1.29) | 1.00 | 1.30 (0.90-1.87) | 2.63 (1.89-3.66) | – |
| Stage IV, HER2+f,g | – | 0.48 (0.33-0.71) | 0.66 (0.51-0.85) | 1.00 | 1.07 (0.86-1.33) | 1.63 (1.30-2.05) | – |
| Stage I, triple negativeh,i | – | 1.21 (0.45-3.24) | 1.10 (0.55-2.19) | 1.00 | 1.78 (1.01-3.14) | 2.22 (1.22-4.04) | – |
| Stage II, triple negativeh,i | – | 1.26 (0.92-1.73) | 0.97 (0.75-1.27) | 1.00 | 1.23 (0.94-1.62) | 2.17 (1.69-2.79) | – |
| Stage III, triple negativeh,i | – | 1.11 (0.85-1.45) | 0.91 (0.73-1.14) | 1.00 | 0.86 (0.67-1.10) | 1.58 (1.26-1.97) | – |
| Stage IV, triple negativeh,i | – | 0.93 (0.69-1.25) | 0.99 (0.78-1.26) | 1.00 | 1.24 (0.99-1.55) | 1.07 (0.83-1.39) | – |
Adjusted for stage, age, year of diagnosis, region, race/ethnicity, marital status, grade, hormone receptor (HR) status, human epidermal growth factor receptor 2 (HER2) status
Adjusted for age, year, region, race/ethnicity, marital status, grade, HR, HER2
Adjusted for age, year, region, race/ethnicity, marital status, grade, HER2
HR+ =estrogen receptor (ER)-positive or progesterone receptor (PR)-positive disease, regardless of HER2 status
HR− =ER-negative and PR-negative disease, regardless of HER2
Adjusted for age, year, region, race/ethnicity, marital status, grade, HR
HER2+ =HER2-positive disease, regardless of ER
Adjusted for age, year, region, race/ethnicity, marital status, grade
Triple negative =HR- and HER2- (2010-2012 diagnoses)
Bolded results are significant
Figure 2.
Scatterplot of the adjusted hazard ratios and 95% confidence intervals in each stage relative to 55-64 years (reference): black=stage I, red=stage II, green=stage III, blue stage IV
The adjusted hazard ratios for breast cancer death for by age for each of the 16 additional cohorts are shown in Table 2 (reference group=ages 55-64 for all analyses except HER2-positive and triple negative cohorts where reference group=ages 51-60). Among those with HR-positive and HR-negative disease, ages 75-84 and ≥85 were consistently associated with higher hazard of breast cancer death regardless of stage, and was most notable for the earlier stage cohorts (e.g. adjusted hazard ratio for age ≥85=3.59 [95% CI=3.12-4.13] for stage I, HR-positive disease and 2.29 [95% CI=2.09-2.52] for those with stage II, HR-positive disease). Among those with HR-negative disease, the adjusted hazard ratios for breast cancer deaths for women age ≥85 ranged from 1.37-1.94 (all with p-values <0.05) with analogous findings for those ages 75-84. Similar findings by age were also observed for those diagnosed in 2010-2012, where age >70 was associated with higher hazard of breast cancer deaths for patients with HER2-positive and triple negative disease, with exception of the stage IV triple negative cohort.
DISCUSSION
In this analysis of nearly 500,000 women with invasive breast cancer living in SEER areas during 2000-2012, we observed worse breast cancer-specific survival for women in the oldest age groups for every stage and subtype analyzed except stage IV triple negative disease (where sample sizes were small), including those with early-stage, advanced stage, and even lower-risk, HR-positive disease. Our results were significant even after adjustment for other relevant factors such as race/ethnicity, marital status, and disease characteristics. Our findings are also consistent with other studies that have documented worse outcomes in older patients across all stages, in the setting of early-stage disease, and among selected women with HR-positive disease.2, 4–6, 10 Although the overall percent of women with early-stage disease dying of breast-cancer was low (1.9% of stage I disease), we observed a more than doubling of this risk for older patients. Further, 39.1% of the oldest patients with stage III disease died as a result of breast cancer compared with 26.6% overall. Our results are sobering given the increasing numbers of older patients anticipated to develop breast cancer within our aging U.S. population in the coming years.13
The reasons for worse outcomes in older patients remain unclear but are likely multifactorial. One important factor is differences in treatment receipt such as under-treatment or lower treatment intensity. It is well established that receipt of adjuvant chemotherapy, trastuzumab, and hormonal therapy reduces risk of recurrence and death across all age groups,14–17 yet multiple studies document sub-optimal systemic treatment and adherence for older patients, including omission of efficacious treatments, receipt of lower intensity and/or non-guideline treatment, or poor adherence to hormonal therapy.2, 18–24 Although evidence supports the safe omission of local therapies such as radiation,25, 26 breast surgery,27 and nodal surgery28–30 without a compromise in survival in the setting of lower risk, HR-positive disease, no study has demonstrated the safety of routine omission of systemic therapies for older patients (and in particular hormonal therapy) except in the lowest-risk cancers.31 It is possible that the larger observed age-related differences in outcomes among those with earlier-stage cancers in our study occurred because the underuse of adjuvant systemic treatment has a disproportionate impact on these patients compared to those with metastatic disease, though this hypothesis cannot be confirmed in our study.
Undoubtedly, under-treatment of breast cancer in older patients may be appropriate in some cases where competing risks and toxicity concerns take priority, but there are likely many situations where older patients will benefit from a more optimized treatment approach. Important work is currently underway to better establish one’s functional age and predict risk for toxicity, all of which will better inform treatment decisions in the near future.32–34 Patient preferences may also contribute to under treatment, which would be appropriate if the care reflects values and preferences for informed patients. Nevertheless, the lack of prospective clinical trial data, guidelines, or even expert consensus on treatments for older patients and patients with comorbidity creates substantial challenges to informed decision-making among older patients. Evidence suggests that such uncertainty is associated with substantial variability in physicians’ treatment recommendations.35
In addition to treatment disparities, variable disease biology may also contribute to the worse outcomes observed for older patients. Although ‘higher risk’ breast cancer subtypes such as triple-negative and HER2-positive cancers occur less frequently in older (vs. younger) patients,1 differences in biology and the genetic underpinnings of breast cancer are likely still present, even among women with similar disease subtypes and in those with HR-positive, ‘lower-risk’ cancers.36,10, 37 Unfortunately, older patients are largely underrepresented in tissue-based genomic studies,38 resulting in a lack of detailed understanding of the genetic differences in cancers by age and their impact on outcomes.
It is important to acknowledge that our findings may also be partially explained by the presence of lead-time bias, as women between ages 50-70 have higher screening mammography rates than older women,39 leading to more lower-risk cancers diagnosed in women ages 50-70, with a longer ‘lead time’ from diagnosis to death. However, U.S. mammography utilization data from 2010 demonstrate small differences in mammography use over the prior two years for women ages 50-74 compared with those ages >75 (79.7% vs. 73.0%), making it unlikely that differences in screening rates fully explain our findings. Further, our observations for worse breast cancer-specific survival were consistent across stages and subtypes, even in those with earlier-stage triple negative cancers, which are less likely to be mammographically-detected.40
In our large, population-based study, we examined outcomes by age in multiple sub-cohorts, providing new and updated epidemiologic data on how age may impact breast cancer-specific survival across stages and disease subtypes. However, we acknowledge several important limitations. First, we lacked extended follow-up for some patients, and analyses of those with HER2-positive and triple negative disease were limited to those diagnosed in 2010-2012. Second, we had no information on comorbidity, performance status, use of mammography or whether cancers were detected by mammography, and receipt of systemic therapy, nor did we have information on treatment decision-making. Thus, the degree to which under treatment versus adherence contributed to our findings could not be determined. Third, given the lack of recurrence data within the SEER registry, the outcomes for metastatic breast cancer could only be examined among patients with de novo stage IV disease. Fourth, detailed information on important factors such individualized socioeconomic variables are not available in SEER but have been shown to impact outcomes. Finally, we focused on cause of death to understand breast cancer-specific outcomes; however, cause-of-death data are subject to misclassification. Reassuringly, our results showed higher rates of non-breast cancer deaths for older patients at each stage but with an increase in the risk of breast cancer death with increasing stage, consistent with what would be expected.
In conclusion, with an increasing number of older patients expected to develop breast cancer in the coming decades, understanding and addressing disparities for older patients is an urgent clinical and research priority. Future studies in older patients should assess reasons for under-treatment for older patients, as the challenges in treating this population are distinct from those seen in other sub-groups. This will require novel approaches which facilitate more active engagement of frail and non-frail older patients in clinical research.
Supplementary Material
Acknowledgments
Funding:RAF-American Cancer Society (125912-MRSG-14-240-01-CPPB) and Susan G. Komen (CCR CCR14298143). NLK-National Cancer Institute (K24CA181510)
Disclosures:RAF-institutional funding from Eisai, Puma Biotechnology. EPW-Honorarium from Genentech, Tesaro, and Lilly; advisory board fees from LEAP
Footnotes
Author Contributions:Rachel Freedman-conceptualization/data curation/formal analysis/funding acquisition/methodology/writing–original draft/writing-review and editing. Nancy Keating-conceptualization/formal analysis/methodology/writing-review and editing. Nancy Lin-conceptualization/formal analysis/methodology/writing-review and editing. Eric Winer-conceptualization/formal analysis/methodology/writing-review and editing. Ines Vaz-Luis-conceptualization/formal analysis/methodology/writing-review and editing. Joyce Lii-conceptualization/data curation/formal analysis/methodology/software/writing-review and editing. Pedro Exman-formal analysis/methodology/writing-review and editing. William Barry-conceptualization/formal analysis/methodology/supervision/writing-review and editing.
References
- 1.Howlader N, Altekruse SF, Li CI, et al. US incidence of breast cancer subtypes defined by joint hormone receptor and HER2 status. J Natl Cancer Inst. 2014;106 doi: 10.1093/jnci/dju055. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2.Schonberg MA, Marcantonio ER, Li D, Silliman RA, Ngo L, McCarthy EP. Breast cancer among the oldest old: tumor characteristics, treatment choices, and survival. J Clin Oncol. 2010;28:2038–2045. doi: 10.1200/JCO.2009.25.9796. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3.Schonberg MA, Marcantonio ER, Ngo L, Li D, Silliman RA, McCarthy EP. Causes of death and relative survival of older women after a breast cancer diagnosis. J Clin Oncol. 2011;29:1570–1577. doi: 10.1200/JCO.2010.33.0472. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4.Chen HL, Zhou MQ, Tian W, Meng KX, He HF. Effect of Age on Breast Cancer Patient Prognoses: A Population-Based Study Using the SEER 18 Database. PLoS One. 2016;11:e0165409. doi: 10.1371/journal.pone.0165409. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.Partridge AH, Hughes ME, Warner ET, et al. Subtype-Dependent Relationship Between Young Age at Diagnosis and Breast Cancer Survival. J Clin Oncol. 2016;34:3308–3314. doi: 10.1200/JCO.2015.65.8013. [DOI] [PubMed] [Google Scholar]
- 6.Smith BD, Jiang J, McLaughlin SS, et al. Improvement in breast cancer outcomes over time: are older women missing out? J Clin Oncol. 2011;29:4647–4653. doi: 10.1200/JCO.2011.35.8408. [DOI] [PubMed] [Google Scholar]
- 7.Satariano WA, Ragland DR. The effect of comorbidity on 3-year survival of women with primary breast cancer. Ann Intern Med. 1994;120:104–110. doi: 10.7326/0003-4819-120-2-199401150-00002. [DOI] [PubMed] [Google Scholar]
- 8.Yancik R, Wesley MN, Ries LA, Havlik RJ, Edwards BK, Yates JW. Effect of age and comorbidity in postmenopausal breast cancer patients aged 55 years and older. Jama. 2001;285:885–892. doi: 10.1001/jama.285.7.885. [DOI] [PubMed] [Google Scholar]
- 9.American Cancer Society. Breast Cancer Facts and Figures 2015-2016. http://www.cancer.org/acs/groups/content/@research/documents/document/acspc-046381.pdf.
- 10.Shak S, Petkov VI, Miller DP, et al. Breast Cancer Specific Survival in 38, 568 Patients with Node Negative Hormone Receptor Positive Invasive Breast Cancer and Oncotype DX Recurrence Score Results in the SEER Database. Presented at the San Antonio Breast Cancer Research Symposium. 2015 doi: 10.1007/s10549-017-4162-3. [DOI] [PubMed] [Google Scholar]
- 11.National Cancer Institute. Surveillance, Epidemilogy, and End Results Program (SEER) https://seer.cancer.gov/. Accessed July 14, 2017.
- 12.Fine JP, Gray RJ. A Proportional Hazards Model for the Subdistribution of a Competing Risk. J of the American Statistical Association. 1999;94:496–509. [Google Scholar]
- 13.Smith BD, Smith GL, Hurria A, Hortobagyi GN, Buchholz TA. Future of cancer incidence in the United States: burdens upon an aging, changing nation. J Clin Oncol. 2009;27:2758–2765. doi: 10.1200/JCO.2008.20.8983. [DOI] [PubMed] [Google Scholar]
- 14.Elkin EB, Hurria A, Mitra N, Schrag D, Panageas KS. Adjuvant chemotherapy and survival in older women with hormone receptor-negative breast cancer: assessing outcome in a population-based, observational cohort. J Clin Oncol. 2006;24:2757–2764. doi: 10.1200/JCO.2005.03.6053. [DOI] [PubMed] [Google Scholar]
- 15.Muss HB, Woolf S, Berry D, et al. Adjuvant chemotherapy in older and younger women with lymph node-positive breast cancer. Jama. 2005;293:1073–1081. doi: 10.1001/jama.293.9.1073. [DOI] [PubMed] [Google Scholar]
- 16.Giordano SH, Duan Z, Kuo YF, Hortobagyi GN, Goodwin JS. Use and outcomes of adjuvant chemotherapy in older women with breast cancer. J Clin Oncol. 2006;24:2750–2756. doi: 10.1200/JCO.2005.02.3028. [DOI] [PubMed] [Google Scholar]
- 17.Early Breast Cancer Trialists’ Collaborative G. Effects of chemotherapy and hormonal therapy for early breast cancer on recurrence and 15-year survival: an overview of the randomised trials. Lancet. 2005;365:1687–1717. doi: 10.1016/S0140-6736(05)66544-0. [DOI] [PubMed] [Google Scholar]
- 18.Vaz-Luis I, Lin NU, Keating NL, et al. Treatment of early-stage human epidermal growth factor 2-positive cancers among medicare enrollees:age and race strongly associated with non-use of trastuzumab. Breast Cancer Res Treat. 2016;159:151–162. doi: 10.1007/s10549-016-3927-4. [DOI] [PubMed] [Google Scholar]
- 19.Freedman RA, Vaz-Luis I, Barry WT, et al. Patterns of chemotherapy, toxicity, and short-term outcomes for older women receiving adjuvant trastuzumab-based therapy. Breast Cancer Res Treat. 2014;145:491–501. doi: 10.1007/s10549-014-2968-9. [DOI] [PubMed] [Google Scholar]
- 20.Vaz-Luis I, Keating NL, Lin NU, Lii H, Winer EP, Freedman RA. Duration and toxicity of adjuvant trastuzumab in older patients with early-stage breast cancer: a population-based study. J Clin Oncol. 2014;32:927–934. doi: 10.1200/JCO.2013.51.1261. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21.Griggs JJ, Culakova E, Sorbero ME, et al. Social and racial differences in selection of breast cancer adjuvant chemotherapy regimens. J Clin Oncol. 2007;25:2522–2527. doi: 10.1200/JCO.2006.10.2749. [DOI] [PubMed] [Google Scholar]
- 22.Freedman RA, Hughes ME, Ottesen RA, et al. Use of adjuvant trastuzumab in women with human epidermal growth factor receptor 2 (HER2)-positive breast cancer by race/ethnicity and education within the National Comprehensive Cancer Network. Cancer. 2013;119:839–846. doi: 10.1002/cncr.27831. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23.Farias AJ, Du XL. Association Between Out-Of-Pocket Costs, Race/Ethnicity, and Adjuvant Endocrine Therapy Adherence Among Medicare Patients With Breast Cancer. J Clin Oncol. 2017;35:86–95. doi: 10.1200/JCO.2016.68.2807. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 24.Biggers A, Shi Y, Charlson J, et al. Medicare D Subsidies and Racial Disparities in Persistence and Adherence With Hormonal Therapy. J Clin Oncol. 2016;34:4398–4404. doi: 10.1200/JCO.2016.67.3350. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 25.van de Water W, Bastiaannet E, Scholten AN, et al. Breast-conserving surgery with or without radiotherapy in older breast patients with early stage breast cancer: a systematic review and meta-analysis. Ann Surg Oncol. 2014;21:786–794. doi: 10.1245/s10434-013-3374-y. [DOI] [PubMed] [Google Scholar]
- 26.Hughes KS, Schnaper LA, Bellon JR, et al. Lumpectomy plus tamoxifen with or without irradiation in women age 70 years or older with early breast cancer: long-term follow-up of CALGB 9343. J Clin Oncol. 2013;31:2382–2387. doi: 10.1200/JCO.2012.45.2615. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 27.Hind D, Wyld L, Reed MW. Surgery, with or without tamoxifen, vs tamoxifen alone for older women with operable breast cancer: cochrane review. Br J Cancer. 2007;96:1025–1029. doi: 10.1038/sj.bjc.6603600. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 28.Martelli G, Boracchi P, De Palo M, et al. A randomized trial comparing axillary dissection to no axillary dissection in older patients with T1N0 breast cancer: results after 5 years of follow-up. Ann Surg. 2005;242:1–6. doi: 10.1097/01.sla.0000167759.15670.14. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 29.Martelli G, Miceli R, Costa A, et al. Elderly breast cancer patients treated by conservative surgery alone plus adjuvant tamoxifen: fifteen-year results of a prospective study. Cancer. 2007;112:481–488. doi: 10.1002/cncr.23213. [DOI] [PubMed] [Google Scholar]
- 30.International Breast Cancer Study G. Rudenstam CM, Zahrieh D, et al. Randomized trial comparing axillary clearance versus no axillary clearance in older patients with breast cancer: first results of International Breast Cancer Study Group Trial 10-93. J Clin Oncol. 2006;24:337–344. doi: 10.1200/JCO.2005.01.5784. [DOI] [PubMed] [Google Scholar]
- 31.Christiansen P, Bjerre K, Ejlertsen B, et al. Mortality rates among early-stage hormone receptor-positive breast cancer patients: a population-based cohort study in Denmark. J Natl Cancer Inst. 2011;103:1363–1372. doi: 10.1093/jnci/djr299. [DOI] [PubMed] [Google Scholar]
- 32.Hurria A, Togawa K, Mohile SG, et al. Predicting chemotherapy toxicity in older adults with cancer: a prospective multicenter study. J Clin Oncol. 2011;29:3457–3465. doi: 10.1200/JCO.2011.34.7625. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 33.Hurria A, Cirrincione CT, Muss HB, et al. Implementing a geriatric assessment in cooperative group clinical cancer trials: CALGB 360401. J Clin Oncol. 2011;29:1290–1296. doi: 10.1200/JCO.2010.30.6985. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 34.Hurria A, Lichtman SM, Gardes J, et al. Identifying vulnerable older adults with cancer: integrating geriatric assessment into oncology practice. J Am Geriatr Soc. 2007;55:1604–1608. doi: 10.1111/j.1532-5415.2007.01367.x. [DOI] [PubMed] [Google Scholar]
- 35.Keating NL, Landrum MB, Klabunde CN, et al. Adjuvant chemotherapy for stage III colon cancer: do physicians agree about the importance of patient age and comorbidity? J Clin Oncol. 2008;26:2532–2537. doi: 10.1200/JCO.2007.15.9434. [DOI] [PubMed] [Google Scholar]
- 36.Paik S, Tang G, Shak S, et al. Gene expression and benefit of chemotherapy in women with node-negative, estrogen receptor-positive breast cancer. J Clin Oncol. 2006;24:3726–3734. doi: 10.1200/JCO.2005.04.7985. [DOI] [PubMed] [Google Scholar]
- 37.Hassett MJ, Silver SM, Hughes ME, et al. Adoption of gene expression profile testing and association with use of chemotherapy among women with breast cancer. J Clin Oncol. 2012;30:2218–2226. doi: 10.1200/JCO.2011.38.5740. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 38.Sorlie T, Perou CM, Tibshirani R, et al. Gene expression patterns of breast carcinomas distinguish tumor subclasses with clinical implications. Proc Natl Acad Sci U S A. 2001;98:10869–10874. doi: 10.1073/pnas.191367098. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 39.Breast Cancer Screening Among Adult Women–Behavior Risk Factor Surveillance System, United States 2010. Supplements. 2012 Jun 15;61(102):46–50. https://www.cdc.gov/mmwr/preview/mmwrhtml/su6102a8.htm Accessed August 17, 2017. [PubMed] [Google Scholar]
- 40.Lin NU, Vanderplas A, Hughes ME, et al. Clinicopathologic features, patterns of recurrence, and survival among women with triple-negative breast cancer in the National Comprehensive Cancer Network. Cancer. 2012;118:5463–5472. doi: 10.1002/cncr.27581. [DOI] [PMC free article] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.