PMCCPMCCPMCC

Search tips
Search criteria 

Advanced

 
Logo of jcoHomeThis ArticleSearchSubmitASCO JCO Homepage
 
J Clin Oncol. Apr 20, 2010; 28(12): 2038–2045.
Published online Mar 22, 2010. doi:  10.1200/JCO.2009.25.9796
PMCID: PMC2860406
Breast Cancer Among the Oldest Old: Tumor Characteristics, Treatment Choices, and Survival
Mara A. Schonberg, Edward R. Marcantonio, Donglin Li, Rebecca A. Silliman, Long Ngo, and Ellen P. McCarthy
From the Division of General Medicine and Primary Care, Department of Medicine, Harvard Medical School, Beth Israel Deaconess Medical Center; Geriatrics Section, Boston University Schools of Medicine and Public Health, Boston University Medical Center, Boston, MA.
Corresponding author: Mara A. Schonberg, MD, MPH, Beth Israel Deaconess Medical Center, 1309 Beacon, Office 202, Brookline, MA 02446; e-mail: mschonbe/at/bidmc.harvard.edu.
Received September 2, 2009; Received December 4, 2009
Purpose
Few data are available on breast cancer characteristics, treatment, and survival for women age 80 years or older.
Patients and Methods
We used the linked Surveillance, Epidemiology and End Results-Medicare data set from 1992 to 2003 to examine tumor characteristics, treatments (mastectomy, breast-conserving surgery [BCS] with radiation therapy or alone, or no surgery), and outcomes of women age 80 years or older (80 to 84, 85 to 89, ≥ 90 years) with stage I/II breast cancer compared with younger women (age 67 to 79 years). We used Cox proportional hazard models to examine the impact of age on breast cancer–related and other causes of death. Analyses were performed within stage, adjusted for tumor and sociodemographic characteristics, treatments received, and comorbidities.
Results
In total, 49,616 women age 67 years or older with stage I/II disease were included. Tumor characteristics (grade, hormone receptivity) were similar across age groups. Treatment with BCS alone increased with age, especially after age 80. The risk of dying from breast cancer increased with age, significantly after age 80. For stage I disease, the adjusted hazard ratio of dying from breast cancer for women age ≥ 90 years compared with women age 67 to 69 years was 2.6 (range, 2.0 to 3.4). Types of treatments received were significantly associated with age and comorbidity, with age as the stronger predictor (26% of women age ≥ 80 years without comorbidity received BCS alone or no surgery compared with 6% of women age 67 to 79 years).
Conclusion
Women age ≥ 80 years have breast cancer characteristics similar to those of younger women yet receive less aggressive treatment and experience higher mortality from early-stage breast cancer. Future studies should focus on identifying tumor and patient characteristics to help target treatments to the oldest women most likely to benefit.
Women age 80 years and older are the fastest growing segment of the US population, and breast cancer is relatively common among these women with nearly 400 cases per 100,000 women.1 Despite the high incidence, little is known about breast cancer characteristics, treatment choices, and survival among the oldest women. These data are important for decision making around breast cancer detection and care. Few randomized controlled trials that evaluated the effectiveness of breast cancer treatments included women age 80 years or older, and most observational studies were limited by small sample sizes in this age range.24 Studies using the Surveillance, Epidemiology and End Results (SEER) -Medicare data set have examined the effectiveness of radiation treatment and chemotherapy among older women with breast cancer.58 However, these studies consistently exclude women with missing stage and those without a histologic diagnosis and/or known hormonal receptivity. Meanwhile, women age 80 years or older are most likely to fall into these categories.2,4,912 Missing data may be one reason that studies differ on whether elderly women present with higher stage disease but with less aggressive tumors than younger women.2,13,14
Although studies consistently show that older women are undertreated for breast cancer, the impact of undertreatment on breast cancer survival among older women remains controversial. Gadjos et al3 found that rates of recurrence were not increased when undertreated women (older than 70 years) were compared with conventionally treated patients, while others have found that undertreatment is associated with recurrence and decreased survival.2,15,16
The purpose of this study was to examine variations in breast cancer tumor characteristics, initial treatments received, and survival among women age 80 to 84, 85 to 89, and ≥ 90 years with early-stage (stage I or II) breast cancer compared with younger women (age 67 to 79 years).
We used data from the National Cancer Institute's linked SEER-Medicare data set. Since 1992, SEER has included 11 population-based tumor registries in the metropolitan areas of San Francisco/Oakland, Detroit, Atlanta, and Seattle; Los Angeles County; the San Jose–Monterey area; and the states of Connecticut, Iowa, New Mexico, Utah, and Hawaii.17 These areas cover approximately 14% of the US population.18 We identified all women age 67 years or older newly diagnosed with breast cancer or ductal carcinoma in situ between 1992 and 2003, excluding those diagnosed on death certificate or at autopsy (n = 102,184). We then excluded women diagnosed with a second cancer within 12 months after their primary breast cancer diagnosis because health care claims cannot reliably discriminate between procedures performed for the index cancer versus the second cancer (100,404 remaining). We further excluded women who had Medicare-managed care insurance within 2 years before through 1 year after diagnosis because their claims data are incomplete (75,286 remaining), and we further excluded women with gaps in their Medicare coverage (66,951 remaining).
We then considered excluding women missing data on American Joint Committee on Cancer (AJCC) Staging 3rd edition (13.5% overall or 19% of women age 80 to 89 years and 39.5% of women age ≥ 90 years), women with a history of non-breast cancer (8.7%), and women without a histologic diagnosis (3.1%). However, when we considered these three exclusions together, 21.9% of our sample would be excluded, including 48.7% of women age ≥ 90 years (Appendix Table A1, online only). Since our study focused on breast cancer characteristics among the oldest women, we chose not to make these exclusions. Instead, we constructed an algorithm using data available on tumor characteristics (extent, number of positive lymph nodes, lymph node invasion, tumor size, and histology) to impute stage for women with missing stage (Appendix A, online only). After imputing stage, only 1.4% of women were still missing stage data. For cases with both known and imputed (AJCC 3rd edition) stage, we tested the extent to which the two measures agreed beyond chance using a two-sided κ statistic. Because of the observed excellent agreement (the two-sided κ statistic was 0.997), we used imputed stage for those women with missing stage in our primary analyses. Our sample included 49,616 women with known or imputed stage I or II disease.
Tumor Characteristics
From SEER, we obtained data on tumor size, regional lymph node involvement, tumor grade, histology, estrogen receptivity (ER), and progesterone receptivity. Variable definitions can be found in Appendix B (online only).
Treatments
We used data from both SEER and Medicare claims to classify initial treatment with surgery or radiation therapy (RT). We considered women to have received surgical and/or RT if reported in SEER or if there were Medicare claims for these treatments within 12 months following diagnosis. We categorized initial treatment as mastectomy, breast-conserving surgery (BCS) plus RT, BCS alone, or no initial surgery. We used Medicare claims to identify receipt of chemotherapy within 12 months following diagnosis among women with ER-positive and lymph node–positive disease, since this population is thought to derive benefit.6,7,19 Women were classified as having any claim versus no claim for chemotherapy (Appendix B).
Survival Outcomes
Survival time was measured from the patient's date of diagnosis until death or December 31, 2005, whichever came first. SEER tracks vital status annually, and death certificates are used to capture underlying cause of death (Appendix B).
Covariates
Patient characteristics included race/ethnicity, marital status, SEER registry, metropolitan versus nonmetropolitan residence, and year of diagnosis.68 Because SEER-Medicare data do not provide individual-level data on income and/or education, we used census tract data and substituted ZIP code–level data when census tract data were not available.20 We grouped median household income and percentage of adults with less than a high school education into quintiles within registry. We defined comorbidity using Klabunde's modification of the Charlson comorbidity index (CCI).21
Analyses
We examined sociodemographic and tumor characteristics by age at diagnosis (67 to 69, 70 to 74, 75 to 79, 80 to 84, 85 to 89, and ≥ 90 years), and we examined receipt of treatment by age at diagnosis and stage using the Mantel-Haenszel test of trend. Because of the large sample size, we knew a priori that even small differences in characteristics among age groups would achieve statistical significance; however, we were most interested in trends by age. We additionally examined the proportion of women who were treated with BCS and RT or mastectomy since these treatments are considered standard and equally effective for early-stage breast cancer.22 Using multinomial logistic regression, we examined the effect of age and comorbidity on receipt of treatments for stage I and II disease separately, adjusting for sociodemographics, tumor characteristics, and year of diagnosis. Since few women did not undergo surgery (n = 843; 1.7%), we did not include “no initial surgery” as a treatment option in these analyses. A category of “missing” was included for each covariate in the models.
To determine the impact of age at diagnosis on breast cancer death, we conducted multivariable Cox proportional hazards regression adjusting for sociodemographic and tumor characteristics, year of diagnosis, initial treatments received, and comorbidity. We censored observations of women alive when follow-up ended. We further tested for interactions between treatment and age at diagnosis on breast cancer mortality. To test for residual confounding, we examined the impact of initial treatment on non-breast cancer survival and overall survival.23 We present the results for non-breast cancer survival since results were similar. We further examined the impact of chemotherapy on breast cancer survival for the subset of women with ER-negative and lymph node–positive tumors adjusting for all covariates.
We performed sensitivity analyses to examine the robustness of our findings. First, we reassessed the impact of treatment on breast cancer survival with propensity score methods, using the “Greedy” match SAS macro24 to minimize bias related to the nonrandom assignment of treatment. Next, we limited our sample to women with known AJCC stage, known histology, and those without a history of cancer. All statistical analyses used SAS version 9.1 (SAS Institute, Cary, NC). The institutional review board approved this study.
Sample Characteristics
Of the 49,616 women included in our final study population, 28,897 had stage I and 16,582 had stage II disease. We used imputed stage for 6,571 women (13.2%). Table 1 displays sociodemographic and tumor characteristics by age at diagnosis. Women age ≥ 80 years (21.8%) were more likely than women age 67 to 79 years (15.0%) to have a CCI of 2+.
Table 1.
Table 1.
Sample and Tumor Characteristics of Older Women With Stage I and II Breast Cancer
Tumor Characteristics
Among women with known tumor grade and progesterone receptivity status, there were no statistically significant differences by age. Differences in ER positivity by age were small (< 5% differences among those with known receptor status). Tumor size increased with age, and the increase was more dramatic after age 80 years. Similarly, the number of women who did not have lymph nodes examined increased with age, and the rate of increase was more dramatic after age 80. Among women who had their lymph nodes examined, those age ≥ 80 years were disproportionately more likely to have positive nodes detected than women age 67 to 79 years.
Initial Treatments
Nearly all women (98.3%) received some surgery for early-stage breast cancer. For women with stage I disease, treatment with BCS + RT declined with age, particularly after age 80 (Fig 1). Mastectomy was the most common treatment among women age 80 to 84 years. Almost all (91.7%) women age 67 to 79 years with stage I disease received mastectomy or BCS + RT compared with 66.8% of women age ≥ 80 years. Among women with stage II disease, mastectomy was the most common treatment regardless of age (Fig 2). However, BCS + RT declined with age. Nearly all women age 66 to 79 years with stage II disease (94.5%) received BCS + RT or mastectomy compared with 76.1% of women age ≥ 80 years. Among women with ER-negative, lymph node–positive disease, receipt of chemotherapy declined significantly with age (Fig 3).
Fig 1.
Fig 1.
Initial treatment for stage I breast cancer by age at diagnosis. Receipt of treatment differed by age at diagnosis for each stage using the Mantel-Haenszel test of trend (P < .001). BCS, breast-conserving surgery; XRT, radiation therapy.
Fig 2.
Fig 2.
Initial treatment for stage II breast cancer by age at diagnosis. Receipt of treatment differed by age at diagnosis for each stage using the Mantel-Haenszel test of trend (P < .001). BCS, breast-conserving surgery; XRT, radiation therapy.
Fig 3.
Fig 3.
Receipt of chemotherapy among women with estrogen receptor–negative, lymph node–positive, stage I/II breast cancer.
In multinomial logistic regression, women age ≥ 80 years were significantly more likely than women age 67 to 79 years to be treated with mastectomy (odds ratio [OR], 2.1; 95% CI, 2.0 to 2.2) or with BCS alone (OR, 4.2; 95% CI, 4.0 to 4.6) compared with BCS + RT. Women with a CCI of 2+ were more likely to receive mastectomy (OR, 1.3; 95% CI, 1.3 to 1.4) or BCS alone (OR, 1.6; 95% CI, 1.5 to 1.8) than BCS + RT. The effect of age was stronger than the effect of comorbidity on receipt of treatment. Among the subset of women with a CCI of 0, 25.8% of women age ≥ 80 years received BCS alone or no surgery compared with 6.0% of women age 65 to 79 years. No significant interactions were observed between age and comorbidity on types of treatment received.
Survival
Overall, median follow-up time was 5.6 years (interquartile range, 3.3 to 8.7). Few women with stage I (4.5%) or stage II (16.1%) disease died of breast cancer. Among women who died, the proportion who died of breast cancer relative to other causes declined with advancing age (Table 1). However, the risk of dying from breast cancer increased significantly with age for women age ≥ 80 years compared with younger women (Table 2). The risk of dying from other causes was greater than the risk of dying from breast cancer at all ages and stages.
Table 2.
Table 2.
HR of Death Due to Breast Cancer and Other Causes by Age at Diagnosis and Stage
Women treated with mastectomy, BCS alone, or no surgery experienced worse breast cancer survival than those treated with BCS + RT (Appendix Table A1). However, for women treated with either mastectomy or BCS alone, the risk of dying from breast cancer was similar to the risk of dying from other causes. Women who received no surgery had a substantially increased risk of dying from breast cancer compared with those treated with BCS + RT, and this risk exceeded their risk of dying from other causes. However, few women did not receive surgery. No significant interactions were observed between age and treatment on breast cancer mortality; however, interactions between age (67 to 79 v ≥ 80 years) and types of treatment were significant for non-breast cancer mortality.
Among women with ER-negative, lymph node–positive breast tumors, we found that chemotherapy reduced breast cancer mortality (adjusted hazard ratio [aHR], 0.8; range, 0.6 to 0.96). Since the interaction of age and chemotherapy was significant (P = .03), we performed subgroup analyses. Chemotherapy was associated with a significant reduction in mortality for women age 67 to 79 years (aHR, 0.6; range, 0.5 to 0.8) and an increased risk of mortality for women age ≥ 80 years (aHR, 1.5; range, 0.9 to 2.3) that did not achieve statistical significance. Chemotherapy was associated with improved non-breast cancer survival among all women (aHR, 0.6; range, 0.4 to 0.8).
Sensitivity Analyses
In analyses of women with known AJCC stage, known histologic diagnosis, and no history of cancer, the impact of age, comorbidity, and treatment on survival were similar (data not shown). Overall, our results for the associations between treatment and survival outcomes were also similar using propensity score methods (Table 3).
Table 3.
Table 3.
HR for Different Breast Cancer Treatments on Breast Cancer Death and Non-Breast Cancer Death
Breast cancer characteristics (eg, tumor grade, histology, hormone receptivity) appear to be similar between women age ≥ 80 years and younger women. However, women age ≥ 80 years receive less aggressive treatment than younger women. Greater comorbidity likely accounts for some of the observed difference; however, among women with a Charlson score of 0, 26% of those age ≥ 80 years did not receive standard treatments (mastectomy or BCS + RT) for early-stage breast cancer compared with only 6% of younger women. We also found that the risk of dying from breast cancer increases significantly after age 80. Our findings suggest that we may be able to identify a subgroup of women age ≥ 80 years who may benefit from more aggressive work-up and treatment of their early-stage breast cancer. Conversely, we may also be able to identify a population of older women on the basis of tumor characteristics, comorbid diseases, and life expectancy who may not need as aggressive treatment. The majority of older women with early-stage disease died from other causes. Future studies are needed to develop tools that can help clinicians appropriately target breast cancer treatments to the oldest women most likely to benefit.
Despite prevailing opinion that breast cancer tumor characteristics are more favorable among older women than younger women, we generally did not find clinically important differences by age at diagnosis for most tumor characteristics. However, the youngest women in our study were older than most women included in other studies.25 It is possible that tumors present with more favorable characteristics with older age but beyond age 67 years, these differences are negligible. Other studies have also failed to show increases in hormone receptor positivity among women age 70 years and older.14,26 Although we and others27 have found that the proportion of women with positive lymph nodes increased with age, we also found that the proportion of women who had their lymph nodes examined declined substantially with age, which may reflect biased sampling. Clinicians may be choosing to sample only lymph nodes of older women who they suspect will be positive.
Regardless of age, we found that the majority of older women undergo surgery for treatment of breast cancer. Among women with stage I disease, BCS + RT is the most common treatment for women age 67 to 79 years. Mastectomy is the most common treatment for women age 80 to 84 years, which may reflect physicians' attempts to treat older women effectively but without radiation. After age 85, BCS alone is the most common treatment. Among women with stage II disease, mastectomy is the most common treatment for all women, regardless of age; however, BCS alone becomes substantially more common after age 80. Some of the oldest women may be undertreated, while others may be being treated appropriately. Future work should focus on identifying tumor and patient characteristics associated with an improved response to aggressive therapy among the oldest women.
As for the impact of RT on older women's breast cancer survival, we found that older women treated with BCS + RT had the best breast cancer survival. However, these women also had the best overall survival, suggesting that unmeasured factors related to survival affected treatment decisions. Clinical trials show that RT after BCS compared with BCS alone reduces breast cancer recurrence among older women with early-stage disease but does not affect survival.8,28,29 Since we found that breast cancer mortality increases significantly after age 80 and these women are the least likely to be treated aggressively, our findings suggest that some older women in good health may benefit from more aggressive treatment.
We found that treatment with chemotherapy was associated with a survival benefit for women age 67 to 79 years with ER-negative, lymph node–positive disease, results similar to those in other studies.6,7 However, chemotherapy tended to be associated with worse breast cancer survival among women age ≥ 80 years. Since few women age ≥ 80 years received chemotherapy, our findings suggest that chemotherapy is reserved for the oldest women with the worst tumor characteristics.
This study has several important limitations. Since this is an observational study, there is potential for selection bias and residual confounding by factors for which we do not have data, such as performance status, social support, and treatment with hormonal therapy. In post hoc sensitivity analyses, we examined the effect of an unmeasured confounder such as hormonal therapy on our estimated aHRs. Assuming that treatment with tamoxifen is more common among women age ≥ 80 years than among younger women30 and that the survival benefit of tamoxifen ranges from 10% to 50% reduction in breast cancer mortality,31 we found that our aHRs would decrease by less than 10% if we were able to adjust for tamoxifen use.32,33 Completion of death certificate data could also differ by age. However, studies have found that coding of cancer on death certificates is accurate, particularly coding of breast cancer death.34,35 In addition, administrative data may underestimate the prevalence of many chronic conditions. Moreover, we needed to exclude women who had missing claims data, the majority of whom had health maintenance organization coverage. Health maintenance organizations tend to include younger and healthier women, which may mean that our sample of women age 67 to 79 years may be older and in poorer health than the overall population. However, this would bias our comparisons between the oldest-old and younger-old toward the null. AJCC staging was modified in 2003 such that women with four or more positive lymph nodes are now classified as stage III. However, only 4.7% of women in our sample had four or more positive nodes. Changes in staging had no effect on women classified as stage I. Finally, although socioeconomic status data were community level, studies have demonstrated moderate associations between individual and aggregate socioeconomic characteristics.20
In summary, breast cancer characteristics are similar among women age ≥ 80 years and younger women. However, women age ≥ 80 years receive less aggressive treatment and are more likely to die from breast cancer. Future studies should focus on identifying tumor and patient characteristics that can be used to help target breast cancer treatments to the oldest women most likely to benefit.
Appendix
Appendix A. Algorithm Used to Impute Stage for Women With Missing Stage
We attempted to use tumor characteristics to define stage for those with known American Joint Committee on Cancer (AJCC) 3rd edition stage from Surveillance, Epidemiology and End Results (SEER) data to create an algorithm to define stage in women with unknown AJCC 3rd edition stage, in the following order:
1. We categorized women with “in situ disease” as noted by the variable extent of disease as stage 0.
2. We categorized women with “metastatic disease” as noted by the variable extent of disease as stage IV.
3. We categorized women with “fixed/matted ipsilateral axillary nodes” and women with “internal mammary node(s), ipsilateral” as defined by the variable lymph node invasion as stage III.
4. We categorized women with “cervical, not otherwise specified; contralateral/bilateral axillary and/or internal mammary; supraclavicular (transverse cervical) other than above” as defined by the variable lymph node invasion as stage IV.
5. We categorized women with “extensive skin involvement, invasion of the chest wall, and/or inflammatory carcinoma” as noted by the variable extent of disease as stage III.
6. We categorized women with tumors > 5 cm in size but with no positive lymph nodes and no lymph node invasion as stage II.
7. We categorized women with tumors > 2 cm but ≤ 5 cm with lymph node involvement or a positive number of lymph nodes as stage II.
8. We categorized women with tumors < 2 cm with lymph node involvement or positive lymph nodes as stage II.
9. We categorized women with “further contiguous extension of their disease” as noted by the variable extent of disease as stage IV.
10. We categorized women with tumors > 5 cm as stage III.
11. We categorized women with tumors > 2 cm but < 5 cm as stage II.
12. We categorized women with tumors < 2 cm as stage I.
13. We categorized women with “tumor extension in their muscle” as noted by the variable extent of disease as stage III.
14. We categorized women who had lymph node invasion or positive lymph nodes but no data on tumor size as stage III.
15. We categorized women with disease “confined to breast” by the variable extent of disease as stage I.
16. We categorized women with “invaded the subcutaneous tissue” as noted by the variable extent of disease as stage II.
17. We categorized women whose histology was reported as “inflammatory” by the variable histology as stage III.
18. We categorized women whose histology was reported “in situ” by the variable histology as stage 0.
Appendix B. Patient Characteristics (Data Source) and Definition
1. Age (Medicare).
Age at diagnosis was entered categorically (67 to 69, 70 to 74, 75 to 79, 80 to 84, 85 to 89, ≥ 90 years). Date of birth and date of diagnosis were assumed to occur on the first of the month, because only month and year were available for these variables.
2. Race/ethnicity (SEER).
Non-Hispanic white, non-Hispanic black, Hispanic, Asian/Pacific Islander, and other. Obtained by registrars from medical records and supplemented by Hispanic surname match.
3. Year of diagnosis (SEER).
Year of diagnosis was entered categorically (1992 to 1995, 1996 to 1999, 2000 to 2003).
4. Marital status (SEER).
Married versus widowed/single/separated/divorced. Unknowns were treated as a separate stratum.
5. Geographic residence (SEER).
SEER tumor registry catchment areas.
6. Metropolitan versus nonmetropolitan residence (SEER).
Big metropolitan/metropolitan/urban versus less urban/rural.
7. Income.
Median household income for the patient's census tract and ZIP code was available from a linkage with the 1990 and 2000 census data. ZIP code–level data were used when census tract data were unavailable. Median income was not available for 255 patients who were treated as a separate stratum. Quintiles of median household income were entered into the final model.
8. Education.
Percentage of adults within census tract with < 12 years of education from the 1990 census and 2000 census. ZIP code–level data were used when census tract data were unavailable. These data were not available for 254 patients who were treated as a separate stratum. Quintiles of proportions with < 12 years of education were entered into the final model.
9. Comorbidity index (Medicare).
We defined comorbidity using Klabunde's modification of the Charlson comorbidity index derived from both inpatient and outpatient claims.21
Tumor Characteristics
1. Size (SEER).
Tumor size was entered continuously into the final model.
2. Grade (SEER).
Well differentiated, moderately differentiated, poorly or undifferentiated, or unknown.
3. Histology (SEER).
Ductal, lobular, mucinous, other, unknown.
4. Estrogen receptor status (SEER).
Positive, borderline, negative, or unknown.
5. Progesterone receptor status (SEER).
Positive, borderline, negative, or unknown.
6. Lymph nodes assessment (SEER).
Pathologic assessment if at least one lymph node was sampled; otherwise, clinical assessment (no lymph nodes examined, no positive lymph nodes, one to three positive lymph nodes and lymph nodes positive, not otherwise specified, four or more positive lymph nodes, or unknown).
Treatments
1. Surgery (SEER and Medicare).
From site-specific surgery variable in SEER and from Medicare claims, breast-conserving surgery (International Classification of Diseases 9th Revision [ICD-9] procedure codes 85.20, 85.21, 85.22, 85.23, or 85.25; Current Procedural Terminology [CPT] codes 19110, 19112, 19120, 19125, 19126, 19160, or 19162) and mastectomy (ICD-9 Procedure Codes 85.41, 85.42, 85.43, 85.44, 85.45, 85.46, 85.47, or 85.48; CPT Codes 19180, 19182, 19200, 19220, or 19240). The most extensive surgical procedure reported by SEER or Medicare within 12 months of diagnosis was considered the definitive surgery. We defined breast-conserving surgery as segmental mastectomy, lumpectomy, quadrantectomy, tylectomy, wedge resection, nipple resection, excisional biopsy, or partial mastectomy that was not otherwise specified.
2. Radiation therapy (SEER and Medicare).
From radiation therapy variable in SEER and from Medicare claims: ICD-9 Procedure Codes 92.21 to 92.29 or 92.29; ICD-9 Diagnosis Codes V58.0, V66.1, or V67.1; CPT codes 77401 to 77525 or 77750 to 77799; and Revenue Center Codes 0330 or 0333.8,17 Patients were considered to have received radiation therapy if either SEER or Medicare reported treatment with radiation therapy within 12 months of diagnosis.
3. Chemotherapy (Medicare).
Chemotherapy use was determined using ICD-9-CM (ICD-9-Clinical Modification) diagnosis and procedure codes, Health Care Common Procedural Coding System (HCPCS) and Revenue Center Codes.6,7,19 ICD-9 Procedure code 99.25; ICD-9 Diagnosis Codes V58.1, V66.2, or V67.2; CPT Codes 96400 to 96549; HCPCS Codes J9000-J9999 or Q0083-Q0085 or J8530, J8600, J8610, J8999, J8510, J8520, J8521; and Revenue Center Codes 0331, 0332, or 0335.8,17 Patients were considered to have received chemotherapy if any of these codes appeared within 12 months of diagnosis.
Outcomes
Death (SEER).
Vital status is tracked annually by using active and passive methods including linkages with state and national death indices, Medicare files, driver's license registration files, voter registration files, Social Security Administration files, national credit agency records, and other databases, as well as contact with patients' hospitals and physicians' offices. Death certificates were used to capture underlying cause of death.
Table A1.
Classification of Stage at Diagnosis by Age at Diagnosis Before and After Imputation
CharacteristicPercentage of Patients in Age Group (years)
Overall (n = 66,951)67-69 (n = 10,075)70-74 (n = 18,405)75-79 (n = 16,906)80-84 (n = 11,990)85-89 (n = 6,444)≥ 90 (n = 3,131)
AJCC 3rd edition stage (prior to imputation)*
    In situ11.815.213.911.910.17.03.6
    I36.139.940.138.333.726.616.4
    II22.623.022.722.822.223.020.3
    III4.43.73.64.14.86.07.9
    IV3.43.23.23.43.53.73.2
    Missing21.915.016.619.625.733.848.7
AJCC 3rd edition stage and women with imputed stage
    In situ14.017.015.713.912.510.09.1
    I47.048.249.449.046.742.130.5
    II27.125.725.426.827.630.934.4
    III5.94.84.85.36.58.412.0
    IV4.63.84.14.54.95.76.5
    Missing1.40.60.70.91.82.97.4
*Staging for women with a histologic diagnosis, no history of cancer, and known American Joint Committee on Cancer (AJCC) 3rd edition stage.
Algorithm to impute stage can be found in Appendix A.
Footnotes
See accompanying editorial on page 1975 and article on page 2046
Supported by Grant No. K23AG028584 from the National Institutes of Health, National Institute on Aging, and a Society of General Internal Medicine, Association of Chiefs of General Internal Medicine, Association of Specialty Professors T. Franklin Williams Scholars Award in Geriatrics, and a 2006 Hartford Geriatrics Health Outcomes Research Scholars Award.
Presented in part at the 32nd Annual Meeting of the Society of General Internal Medicine, May 13-16, 2009, Miami Beach, FL, and the 2009 Annual Meeting of the American Geriatrics Society, April 29-May 2, 2009, Chicago, IL.
Authors' disclosures of potential conflicts of interest and author contributions are found at the end of this article.
AUTHORS' DISCLOSURES OF POTENTIAL CONFLICTS OF INTEREST
The author(s) indicated no potential conflicts of interest.
AUTHOR CONTRIBUTIONS
Conception and design: Mara A. Schonberg, Edward R. Marcantonio, Rebecca A. Silliman, Ellen P. McCarthy
Financial support: Mara A. Schonberg
Administrative support: Mara A. Schonberg
Collection and assembly of data: Mara A. Schonberg, Donglin Li
Data analysis and interpretation: Mara A. Schonberg, Edward R. Marcantonio, Donglin Li, Rebecca A. Silliman, Long Ngo, Ellen P. McCarthy
Manuscript writing: Mara A. Schonberg, Edward R. Marcantonio, Donglin Li, Rebecca A. Silliman, Long Ngo, Ellen P. McCarthy
Final approval of manuscript: Mara A. Schonberg, Edward R. Marcantonio, Donglin Li, Rebecca A. Silliman, Long Ngo, Ellen P. McCarthy
1. Breast Cancer Facts and Figures 2007-2008. http://www.cancer.org/downloads/stt/bcff-final.pdf.
2. Bouchardy C, Rapiti E, Fioretta G, et al. Undertreatment strongly decreases prognosis of breast cancer in elderly women. J Clin Oncol. 2003;21:3580–3587. [PubMed]
3. Gajdos C, Tartter PI, Bleiweiss IJ, et al. The consequence of undertreating breast cancer in the elderly. J Am Coll Surg. 2001;192:698–707. [PubMed]
4. Hébert-Croteau N, Brisson J, Latreille J, et al. Compliance with consensus recommendations for the treatment of early stage breast carcinoma in elderly women. Cancer. 1999;85:1104–1113. [PubMed]
5. Du X, Goodwin JS. Increase of chemotherapy use in older women with breast carcinoma from 1991 to 1996. Cancer. 2001;92:730–737. [PMC free article] [PubMed]
6. Elkin EB, Hurria A, Mitra N, et al. 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. [PubMed]
7. Giordano SH, Duan Z, Kuo YF, et al. Use and outcomes of adjuvant chemotherapy in older women with breast cancer. J Clin Oncol. 2006;24:2750–2756. [PubMed]
8. Smith BD, Gross CP, Smith GL, et al. Effectiveness of radiation therapy for older women with early breast cancer. J Natl Cancer Inst. 2006;98:681–690. [PubMed]
9. Mandelblatt JS, Hadley J, Kerner JF, et al. Patterns of breast carcinoma treatment in older women: Patient preference and clinical and physical influences. Cancer. 2000;89:561–573. [PubMed]
10. Ballard-Barbash R, Potosky AL, Harlan LC, et al. Factors associated with surgical and radiation therapy for early stage breast cancer in older women. J Natl Cancer Inst. 1996;88:716–726. [PubMed]
11. Enger SM, Thwin SS, Buist DS, et al. Breast cancer treatment of older women in integrated health care settings. J Clin Oncol. 2006;24:4377–4383. [PMC free article] [PubMed]
12. Litvak DA, Arora R. Treatment of elderly breast cancer patients in a community hospital setting. Arch Surg. 2006;141:985–990. [PubMed]
13. Giordano SH, Hortobagyi GN, Kau SW, et al. Breast cancer treatment guidelines in older women. J Clin Oncol. 2005;23:783–791. [PubMed]
14. Gazetas P, Estabrook A, O'Neil J, et al. Importance of adequate staging and of hormone receptors in women older than age 70 with breast cancer. Ann Surg. 1992;216:22–26. [PubMed]
15. Owusu C, Lash TL, Silliman RA. Effect of undertreatment on the disparity in age-related breast cancer-specific survival among older women. Breast Cancer Res Treat. 2007;102:227–236. [PubMed]
16. Yood MU, Owusu C, Buist DS, et al. Mortality impact of less-than-standard therapy in older breast cancer patients. J Am Coll Surg. 2008;206:66–75. [PubMed]
17. National Cancer Institute. Procedure Codes for SEER-Medicare Analyses. http://healthservices.cancer.gov/seermedicare/considerations/procedure_codes.html.
18. Ries LAG, Kosary CL, Hankey BF, et al. Bethesda, MD: 1997. SEER Cancer Statistics Review, 1973-1994. National Cancer Institute. NIH Publication No. 97-2789.
19. Warren JL, Harlan LC, Fahey A, et al. Utility of the SEER-Medicare data to identify chemotherapy use. Med Care. 2002;40(suppl 8)(IV):55–61. [PubMed]
20. Bach PB, Guadagnoli E, Schrag D, et al. Patient demographic and socioeconomic characteristics in the SEER-Medicare database applications and limitations. Med Care. 2002;40(suppl 8)(IV):19–25. [PubMed]
21. National Cancer Institute. SEER-Medicare: Calculation of Comorbidity Weights. http://healthservices.cancer.gov/seermedicare/program/comorbidity.html.
22. NCCN Clinical Practice Guidelines in Oncology. Breast Cancer. V.I. 2009. http://www.nccn.org/professionals/physician_gls/PDF/breast.pdf.
23. Giordano SH, Kuo YF, Duan Z, et al. Limits of observational data in determining outcomes from cancer therapy. Cancer. 2008;112:2456–2466. [PubMed]
24. Parsons LS. Reducing bias in a propensity score matched-pair sample using greedy matching techniques. SUGI 26. http://www2.sas.com/proceedings/sugi26/p214-26.pdf.
25. Diab SG, Elledge RM, Clark GM. Tumor characteristics and clinical outcome of elderly women with breast cancer. J Natl Cancer Inst. 2000;92:550–556. [PubMed]
26. Fernö M, Borg A, Johansson U, et al. Estrogen and progesterone receptor analyses in more than 4,000 human breast cancer samples: A study with special reference to age at diagnosis and stability of analyses—Southern Swedish Breast Cancer Study Group. Acta Oncol. 1990;29:129–135. [PubMed]
27. Wildiers H, Van Calster B, van de Poll-Franse LV, et al. Relationship between age and axillary lymph node involvement in women with breast cancer. J Clin Oncol. 2009;27:2931–2937. [PubMed]
28. Hughes KS, Schnaper LA, Berry D, et al. Lumpectomy plus tamoxifen with or without irradiation in women 70 years of age or older with early breast cancer. N Engl J Med. 2004;351:971–977. [PubMed]
29. Fyles AW, McCready DR, Manchul LA, et al. Tamoxifen with or without breast irradiation in women 50 years of age or older with early breast cancer. N Engl J Med. 2004;351:963–970. [PubMed]
30. Silliman RA, Guadagnoli E, Rakowski W, et al. Adjuvant tamoxifen prescription in women 65 years and older with primary breast cancer. J Clin Oncol. 2002;20:2680–2688. [PubMed]
31. Owusu C, Lash TL, Silliman RA. Effectiveness of adjuvant tamoxifen therapy among older women with early stage breast cancer. Breast J. 2007;13:374–382. [PubMed]
32. Lin DY, Psaty BM, Kronmal RA. Assessing the sensitivity of regression results to unmeasured confounders in observational studies. Biometrics. 1998;54:948–963. [PubMed]
33. Mitra N, Heitjan DF. Sensitivity of the hazard ratio to nonignorable treatment assignment in an observational study. Stat Med. 2007;26:1398–1414. [PubMed]
34. Kircher T, Nelson J, Burdo H. The autopsy as a measure of accuracy of the death certificate. N Engl J Med. 1985;313:1263–1269. [PubMed]
35. Mollo F, Berry DA, Cirrincione CT, et al. Reliability of death certifications for different types of cancer. An autopsy survey. Pathol Res Pract. 1986;181:442–447. [PubMed]
Articles from Journal of Clinical Oncology are provided here courtesy of
American Society of Clinical Oncology