Survival benefit of adjuvant chemotherapy in elderly patients with stage I triple-negative breast cancer: a cohort study based on the SEER database

Introduction

Triple-negative breast cancer (TNBC), lacking estrogen receptor (ER) and progesterone receptor (PR) expression and human epidermal growth factor receptor 2 (HER2) gene amplification, accounting for 15–20% of all breast cancers, is the most destructive breast cancer with poor prognosis (1-3). TNBC cannot take advantage of targeted therapy and endocrine therapy without valuable therapeutic targets, and postoperative adjuvant chemotherapy (POCT) remains a helpful treatment for TNBC (4). According to current recommendations, most early TNBC patients should receive adjuvant chemotherapy (5). The National Comprehensive Cancer Network (NCCN) adopts adjuvant chemotherapy as standard treatment for all lymph node positive and node-negative TNBC patients with a primary tumor above 1 cm. For TNBC patients with a diameter of less than 1 cm without lymph node metastasis, NCCN guidelines usually do not recommend adjuvant chemotherapy for T1a-TNBC, and for T1b-TNBC, the chemotherapy recommendation is not clear, suggesting that ‘consider’ adjuvant chemotherapy (4,6,7). The Chinese Society of Clinical Oncology (CSCO) guidelines suggest that all TNBC patients should receive POCT without any specific recommendations (6,8).

Around 40% of breast cancers are found in people over 65 years old, which has been an upward trend recently (9). Due to factors such as short life expectancy and many complications, clinical trials are generally excluded from the elderly population. Adjuvant chemotherapy is not widely studied in elderly women with breast cancer, and insufficient data are collected to support definitive chemotherapy recommendations. The effect of POCT in elderly patients with stage I TNBC has become a clinical issue. We explored the correlation between POCT and survival outcomes in stage I TNBC patients ≥65 years old based on large-scale Surveillance, Epidemiology, and End Results (SEER) database, providing a theoretical basis for clinical practice. We present this article in accordance with the STROBE reporting checklist (available at https://tcr.amegroups.com/article/view/10.21037/tcr-23-123/rc).

Methods

Data sources

The data are derived from the SEER database, a publicly available resource for cancer research in the USA. The SEER database gathers cancer cases data since 1973, including 18 cancer registration centers, covering about 28% of the population in the USA. We recruited the SEER*Stat software 8.3.6 to obtain the variables.

Study population

We retrospectively reviewed all information on patients with primary breast cancer from the SEER database between 2010 and 2017, and gathered pertinent pathology, population, and survival data. The inclusion criteria were as follows: (I) female patients aged ≥65 years; (II) TNBC subtype (negative expression of ER, PR, and HER2); (III) diagnosis of tumor stage I based on the 7th edition of the American Joint Committee on Cancer (AJCC) stage (converting 2016–2017 UICC stage to 7th edition AJCC stage); (IV) previous lumpectomy or mastectomy; (V) infiltrating carcinoma as the histology type (International Classification of Disease, 3rd edition, ICD-O-3 code 8500/3, 8522/3 and 8523/3). For better analysis, all quantitative data were converted to categorical variables using the X-tile software.

Statistical analysis

The study population was divided into POCT and non-POCT groups. Our outcomes included overall survival (OS) and breast cancer-specific survival (BCSS). OS was described as the interval from diagnosis to death, and BCSS as the period from diagnosis to the date of death from breast cancer. Univariate analysis was employed to assess risk factors related to chemotherapy (P<0.05), and multivariate stepwise logistic regression models were constructed based on univariate analyses (P<0.05). We carried out propensity score matching (PSM) to eliminate heterogeneity between the two groups. POCT groups and non-POCT groups were matched by nearest neighbor with caliper value of 0.01 in a 1:1 ratio. Kaplan-Meier plots were tested to compare OS and BCSS between POCT and non-POCT groups to determine survival differences, and log-rank tests were used to compare survival distributions. Subgroup analysis of the effects of OS and BCSS on the two groups at different stages was carried out. Z test was performed to compare survival 5-year OS and BCSS rates between stage IA and stage IB tumors. Univariate and multivariate Cox proportional risk regression models were constructed to assess prognostic factors related to survival. All analyses were carried out with R 4.1.2 software, and P values <0.05 was regarded as statistically significant differences. The study was conducted in accordance with the Declaration of Helsinki (as revised in 2013).

Results

Patient baseline characteristics

A total of 3,307 female TNBC patients aged ≥65 years were enrolled in this study, of whom 1,653 (50.0%) received chemotherapy and 1,985 (60.0%) received radiotherapy. The median survival time was 48 (range, 0–107) months. There were 1,281 (38.7%) patients aged 65–69 years, 948 (28.7%) patients aged 70–74 years, 614 (18.6%) patients aged 75–79 years, and 464 (14.0%) patients aged ≥80 years. In terms of the clinical stage of breast cancer, 3,216 (97.2%) patients had stage IA, 91 (2.8%) patients had stage IB. In addition, 1,748 (52.9%) of the patients were married. Their demographic and clinicopathologic characteristics are described in Table 1.

Predictors of POCT

The use of chemotherapy decreased with age in older patients. 68.2% of those aged 65–69, 54.2% of those aged 70–74, 34.0% of those aged 75–79, and 12.1% of those older than 80 received POCT. In multivariate analysis, younger age (P<0.001), higher tumor grade (P<0.001), married (P<0.001), postoperative radiotherapy (P<0.001), lumpectomy (P=0.014), larger tumor size (P<0.001), and recent year of diagnosis (P=0.030) were related to an added likelihood of receiving POCT. Univariate and multivariate logistic regressions of POCT outcomes, demographics and clinical characteristics of the study population are demonstrated in Table 1.

Survival outcomes

Obvious differences were shown between the POCT group and the non-POCT group at baseline for the original data in terms of age (P<0.001), marital status (P<0.001), year of diagnosis (P=0.003), grade (P<0.001), tumor size (P<0.001), type of surgery (P<0.001), and radiotherapy (P<0.001). We used PSM to eliminate the interference and balanced the heterogeneity. After PSM, the baseline data were essentially the same in the two groups (P>0.05), demonstrating that the variables were offsetting. Ultimately, 1,842 patients were enrolled, half of whom received chemotherapy. Their demographic and clinical characteristics before and after PSM treatment are demonstrated in Table 2.

The median follow-up time was 51 (range, 2–107) months and 51 (range, 0–107) months for the POCT and non-POCT groups, respectively. We evaluated OS and BCSS before and after PSM for TNBC patients at stage I using the Kaplan-Meier method. Overall, POCT notably improved OS (before PSM: P<0.0001, after PSM: P<0.0001) (Figure 1) rather than BCSS (before PSM: P=0.42, after PSM: P=0.26) (Figure 1). Before PSM, 5-year OS of the POCT group and the non-POCT group were 90.9% and 82.6%, respectively (P<0.001), while 5-year BCSS was similar, 94.2% and 93.4%, respectively (P=0.42); after PSM, the 5-year OS was 90.0% and 83.0% in the POCT and non-POCT groups, respectively (P<0.001), while the 5-year BCSS was similar in the POCT and non-POCT groups, 94.4% and 92.5%, respectively (P=0.26) (Table 3).

Figure 1 Kaplan-Meier curves comparing OS and BCSS for the POCT versus non-POCT groups in whole cohort. (A) OS before PSM. (B) BCSS before PSM. (C) OS after PSM. (D) BCSS after PSM. POCT, postoperative adjuvant chemotherapy; OS, overall survival; BCSS, breast cancer-specific survival; PSM, propensity score matching.

Subgroup analysis

We conducted a subgroup analysis to determine the impact of prognostic causes. Before PSM, POCT could improve OS (P<0.0001) but not BCSS (P=0.64) in patients with stage IA TNBC (Figure 2A,2B). After PSM, for TNBC patients in stage IA, POCT could improve OS in stage IA (P=0.00035) and could not improve BCSS (P=0.47) (Figure 2C,2D). Before PSM, for TNBC patients in stage IB, POCT could improve OS (P=0.0046) and BCSS (P=0.048) (Figure 3A,3B). After PSM, POCT could improve OS (P=0.0058), the improvement of BCSS was not statistically significant (P=0.053) (Figure 3C,3D).

Figure 2 Kaplan-Meier curves comparing OS and BCSS for the POCT versus non-POCT groups in IA subgroup. (A) OS before PSM. (B) BCSS before PSM. (C) OS after PSM. (D) BCSS after PSM. POCT, postoperative adjuvant chemotherapy; OS, overall survival; BCSS, breast cancer-specific survival; PSM, propensity score matching.

Figure 3 Kaplan-Meier curves comparing OS and BCSS for the POCT versus non-POCT groups in IB subgroup. (A) OS before PSM. (B) BCSS before PSM. (C) OS after PSM. (D) BCSS after PSM. POCT, postoperative adjuvant chemotherapy; OS, overall survival; BCSS, breast cancer-specific survival; PSM, propensity score matching.

Cox proportional hazards models for OS and BCSS

The correlation between clinical characteristics and OS, BCSS was shown using univariate and multifactorial Cox proportional risk regression models (Table 4). After PSM matching, univariate Cox regression showed that age (P<0.001), chemotherapy (P<0.001), marital status (P<0.001), race (P=0.005), radiotherapy (P=0.006), tumor stage (P<0.001), and tumor size (P<0.001) were significant factors affecting TNBC patients’ predictors of OS. Age (P=0.001), race (P<0.001), stage (P<0.001), and tumor size (P<0.001) were notable forecasters of BCSS. Statistically significant variables in the univariate were involved in multifactorial regression model, and multifactorial Cox regression analysis demonstrated that age (P=0.003), chemotherapy (P<0.001), race (P=0.011), radiotherapy (P=0.023), tumor stage (P<0.001), and tumor size (P=0.010) were independent prognostic factors for OS. Age (P=0.019), race (P<0.001), tumor stage (P<0.001), and tumor size (P<0.001) were independent prognostic factors for BCSS.

Discussion

Adjuvant chemotherapy can increase the prognosis of early TNBC patients and prolong their survival time (10-12). SEER database also demonstrates the important role of adjuvant chemotherapy in early stage TNBC (6,13). Meanwhile, the NCCN guidelines state that adjuvant chemotherapy is recommended to be “considered” for stage I (pT1bN0M0) TNBC patients (5). However, only a few elderly patients were enrolled in the above retrospective study. To our knowledge, there is no prospective or retrospective study to evaluate the survival outcomes of POCT for these specific older TNBC patients. Therefore, this study collected the patients of SEER database above 65 years with stage I TNBC from 2010 to 2016 to analyze the correlation between their receipt of POCT and survival outcomes.

The employment of adjuvant chemotherapy is limited in elderly patients. Studies have indicated that the likelihood of older patients with TNBC receiving chemotherapy is minimal, with the main factor being concerns about tolerability in people with reduced functional reserve (14-18). In our study, we observed that the ratio of elderly patients receiving POCT reduced with increasing age (68.2%, 54.2%, 34.0%, and 12.1% for 65–69, 70–74, 75–79, and ≥80 age subgroups, respectively), which is similar to that reported previous studies (16,19,20). Furthermore, our results also indicate that age is an independent predicator affecting survival, which is consistent with previous reports (15). However, some researches indicate that elderly patients with TNBC may own more sound outcomes than younger patients. Providing adjuvant chemotherapy based on age alone may have a negative impact on the prognosis of such patients (21-23). It may be because age-related alterations in TNBC and general breast cancer need to be viewed in the context of the underlying genomic phenotype. Patient age is interrelated with changes in Ki-67 expression, PIK3CA mutations, and luminal androgen receptor subtypes. Therefore, in addition to the age at diagnosis, more consideration should be given to the impact of tumor biology in making treatment decisions (17).

A retrospective study of 177 elderly patients with stage I–III TNBC in Peking Union Medical College Hospital showed that those receiving chemotherapy were younger with more advanced disease. Chemotherapy with more sound BCSS and OS, and the recurrence-free survival (RFS) rate receiving chemotherapy was improved (22). However, the impact of adjuvant chemotherapy on the survival outcome of stage I patients was not specifically assessed here. In this study, it was observed the 5-year OS of the POCT group was pronouncedly improved compared to that of the non-POCT group, both before and after matching. Difference of 5-year BCSS between POCT and non-POCT was not statistical. However, there are studies that are inconsistent with our results. It has been reported that adjuvant chemotherapy does not enhance the prognosis of elderly TNBC (24). Huang et al. suggested that chemotherapy did not improve the survival rate of elderly patients with stage I TNBC, while the situation for patients with stage II and III is the opposite (25).

In addition, we conducted exploratory subgroup analysis to determine the influence of prognostic factors. We divided stage I TNBC into stage IA and stage IB to compare the survival rate between POCT group and non-POCT group. There was an obvious difference between stage IA and stage IB in the benefit of POCT for TNBC patients in OS, but the benefits for BCSS were only significant in stage IB, and the difference in BCSS between the two groups in stage IA was not pronounced, indicating that POCT can decrease the risk of death from all causes or from breast cancer in stage IB, but not in stage IA death from breast cancer.

Limitations also exist here. First, the inherent selection bias is unavoidable in this retrospective analysis. Second, patients in the non-POCT group may be healthier than those in the POCT group. Third, the data for elderly patients were unevenly distributed, with few elderly patients in T1a or T1b tumors, so a more detailed subgroup analysis by stage was not possible. Fourth, some biological characteristics such as Ki-67, BRCA1 and BRCA2 mutations were not registered in the SEER database. Fifth, SEER did not record the whole-body specific treatment information such as adjuvant chemotherapy regimen, dose and number of adjuvant chemotherapy cycles, which may affect the impact of chemotherapy on OS and BCSS. Sixth, a notable shortcoming is that chemotherapy records in the SEER database are categorized as “no/unknown” and “yes”. Data for 921 patients with definite chemotherapy were recorded, but it is not sure that whether the patients recorded as “no/unknown” had in fact not received chemotherapy. This study also has certain advantages. First, as the first study using PSM in elderly patients with stage I TNBC, the results are more dependable than those without PSM. This is the first retrospective study to analyze the effect of POCT on elderly patients with stage I TNBC using a public database. Secondly, the advantage of our research is to focus on the newest obtainable data in the SEER database and strictly exclude lost data to ensure the reliability of our research.

Although our study provides evidence that older women over 65 with stage I TNBC can benefit from POCT, more comprehensive genetic stratification with advanced molecular biology techniques is needed and should be performed in order to identify the appropriate chemotherapy population and decide on adjuvant chemotherapy regimens.

Conclusions

In our study, the use of POCT decreased with age in elderly female patients over 65 years with stage I TNBC, and elderly female patients appeared to benefit from POCT. POCT may be a promising tool for elderly female patients with stage I TNBC in the future.

Acknowledgments

The authors would like to thank the SEER database for providing valuable public datasets.

Funding: This work was supported by the Department of Special Project on Basic Research for Local Undergraduate Universities (Part) in Yunnan Province, China (No. 202101BA070001-121).

Footnote

Reporting Checklist: The authors have completed the STROBE reporting checklist. Available at https://tcr.amegroups.com/article/view/10.21037/tcr-23-123/rc

Peer Review File: Available at https://tcr.amegroups.com/article/view/10.21037/tcr-23-123/prf

Conflicts of Interest: All authors have completed the ICMJE uniform disclosure form (available at https://tcr.amegroups.com/article/view/10.21037/tcr-23-123/coif). The authors have no conflicts of interest to declare.

Ethical Statement: The authors are accountable for all aspects of the work in ensuring that questions related to the accuracy or integrity of any part of the work are appropriately investigated and resolved. The study was conducted in accordance with the Declaration of Helsinki (as revised in 2013).

Open Access Statement: This is an Open Access article distributed in accordance with the Creative Commons Attribution-NonCommercial-NoDerivs 4.0 International License (CC BY-NC-ND 4.0), which permits the non-commercial replication and distribution of the article with the strict proviso that no changes or edits are made and the original work is properly cited (including links to both the formal publication through the relevant DOI and the license). See: https://creativecommons.org/licenses/by-nc-nd/4.0/.

References

1. Bertucci F, Houlgatte R, Benziane A, et al. Gene expression profiling of primary breast carcinomas using arrays of candidate genes. Hum Mol Genet 2000;9:2981-91. [Crossref] [PubMed]
2. Foulkes WD, Smith IE, Reis-Filho JS. Triple-negative breast cancer. N Engl J Med 2010;363:1938-48. [Crossref] [PubMed]
3. Garrido-Castro AC, Lin NU, Polyak K. Insights into Molecular Classifications of Triple-Negative Breast Cancer: Improving Patient Selection for Treatment. Cancer Discov 2019;9:176-98. [Crossref] [PubMed]
4. Fasano GA, Bayard S, Chen Y, et al. Benefit of adjuvant chemotherapy in node-negative T1a versus T1b and T1c triple-negative breast cancer. Breast Cancer Res Treat 2022;192:163-73. [Crossref] [PubMed]
5. Steenbruggen TG, van Werkhoven E, van Ramshorst MS, et al. Adjuvant chemotherapy in small node-negative triple-negative breast cancer. Eur J Cancer 2020;135:66-74. [Crossref] [PubMed]
6. Du ZL, Wang Y, Wang DY, et al. Evaluation of a beneficial effect of adjuvant chemotherapy in patients with stage I triple-negative breast cancer: a population-based study using the SEER 18 database. Breast Cancer Res Treat 2020;183:429-38. [Crossref] [PubMed]
7. Zhai Z, Zheng Y, Yao J, et al. Evaluation of Adjuvant Treatments for T1 N0 M0 Triple-Negative Breast Cancer. JAMA Netw Open 2020;3:e2021881. [Crossref] [PubMed]
8. Gradishar WJ, Moran MS, Abraham J, et al. NCCN Guidelines® Insights: Breast Cancer, Version 4.2021. J Natl Compr Canc Netw 2021;19:484-93. [Crossref] [PubMed]
9. Gennari R, Curigliano G, Rotmensz N, et al. Breast carcinoma in elderly women: features of disease presentation, choice of local and systemic treatments compared with younger postmenopasual patients. Cancer 2004;101:1302-10. [Crossref] [PubMed]
10. Loi S, Drubay D, Adams S, et al. Tumor-Infiltrating Lymphocytes and Prognosis: A Pooled Individual Patient Analysis of Early-Stage Triple-Negative Breast Cancers. J Clin Oncol 2019;37:559-69. [Crossref] [PubMed]
11. Li Y, Zhou Y, Mao F, et al. Adjuvant addition of capecitabine to early-stage triple-negative breast cancer patients receiving standard chemotherapy: a meta-analysis. Breast Cancer Res Treat 2020;179:533-42. [Crossref] [PubMed]
12. Pistelli M, Pagliacci A, Battelli N, et al. Prognostic factors in early-stage triple-negative breast cancer: lessons and limits from clinical practice. Anticancer Res 2013;33:2737-42. [PubMed]
13. Zhang J, Wang W, Wang J, et al. Survival Outcome and Impact of Chemotherapy in T1 Node-Negative Triple-Negative Breast Cancer: A SEER Database Analysis. J Oncol 2020;2020:8880727. [Crossref] [PubMed]
14. Yoon J, Knapp G, Quan ML, et al. Cancer-Specific Outcomes in the Elderly with Triple-Negative Breast Cancer: A Systematic Review. Curr Oncol 2021;28:2337-45. [Crossref] [PubMed]
15. Ward SE, Holmes GR, Ring A, et al. Adjuvant Chemotherapy for Breast Cancer in Older Women: An Analysis of Retrospective English Cancer Registration Data. Clin Oncol (R Coll Radiol) 2019;31:444-52. [Crossref] [PubMed]
16. 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;24:2757-64. [Crossref] [PubMed]
17. Aine M, Boyaci C, Hartman J, et al. Molecular analyses of triple-negative breast cancer in the young and elderly. Breast Cancer Res 2021;23:20. [Crossref] [PubMed]
18. Jauhari Y, Dodwell D, Gannon MR, et al. The influence of age, comorbidity and frailty on treatment with surgery and systemic therapy in older women with operable triple negative breast cancer (TNBC) in England: A population-based cohort study. Eur J Surg Oncol 2021;47:251-60. [Crossref] [PubMed]
19. Kaplan HG, Malmgren JA, Atwood MK. Triple-negative breast cancer in the elderly: Prognosis and treatment. Breast J 2017;23:630-7. [Crossref] [PubMed]
20. Wan S, Jubelirer S. Geographic access and age-related variation in chemotherapy use in elderly with metastatic breast cancer. Breast Cancer Res Treat 2015;149:199-209. [Crossref] [PubMed]
21. Aapro M, Wildiers H. Triple-negative breast cancer in the older population. Ann Oncol 2012;23:vi52-5. [Crossref] [PubMed]
22. Xiu M, Zhang P, Li Q, et al. Chemotherapy Decision-Making and Survival Outcomes in Older Women With Early Triple-Negative Breast Cancer: Evidence From Real-World Practice. Front Oncol 2022;12:867583. [Crossref] [PubMed]
23. Janeva S, Zhang C, Kovács A, et al. Adjuvant chemotherapy and survival in women aged 70 years and older with triple-negative breast cancer: a Swedish population-based propensity score-matched analysis. Lancet Healthy Longev 2020;1:e117-24. [Crossref] [PubMed]
24. Bulut N, Altundag K. Excellent clinical outcome of triple-negative breast cancer in younger and older women. J BUON 2015;20:1276-81. [PubMed]
25. Huang K, Zhang J, Yu Y, et al. The impact of chemotherapy and survival prediction by machine learning in early Elderly Triple Negative Breast Cancer (eTNBC): a population based study from the SEER database. BMC Geriatr 2022;22:268. [Crossref] [PubMed]