Elderly women constitute an important part of breast cancer patients and have different biological and clinical characteristics compared with younger women. The physical decline of elderly patients increases with age, and poor physical status, many accompanying chronic diseases, and poor tolerance to treatment are also notable characteristics of such patients. In addition, due to the lack of prospective studies in elderly patients, the beneficial outcomes of chemotherapy in such patients are still controversial, and the actual practice is still influenced by the subjective opinion of clinicians, who lack reliable evidence to guide their treatment plans. In conclusion, it remains uncertain whether adjuvant chemotherapy translates into survival benefit after 70 years old. Therefore, it is of great significance to develop an individualized-level disease risk assessment model.
Sharon et al, in an observational study based on data collected from the SEER database, used logistic regression analysis to determine factors related to chemotherapy and Cox proportional hazards models to calculate the hazard of death for patients with and without chemotherapy 10. Another study based on the SEER database used propensity score methods and multivariable proportional hazards regression to evaluate the effect of chemotherapy for patients with hormone receptor (HR)-negative breast cancer 11. However, the effects of other tumor variables, such as lymph node status, tumor size and grade, and HER2 expression, were not analyzed, and the subgroup of elderly women most likely to benefit from chemotherapy remains uncertain. In this study, we compared clinicopathological characteristics between the chemotherapy group and the non-chemotherapy group and performed PSM and IPTW matching analyses to ensure that differences in outcomes were not due to demographic or pathological baseline imbalances between the two groups. The risk factors affecting BCSS were screened out by univariate and multivariate Cox regression analyses, and a predictive nomogram was constructed accordingly. It is verified that the model has good predictive performance. Our research can help clinicians accurately screen out patients who can benefit from chemotherapy and provide a reference for clinical treatment.
Whether elderly patients can truly benefit from chemotherapy is beyond doubt. Tamirisa et al found that chemotherapy improved overall survival in elderly breast cancer patients with positive nodes, positive estrogen receptor and multiple comorbidities 12. The EBCTCG meta-analysis shows that patients over 70 years old have improved recurrence-free and overall survival after receiving combination chemotherapy, even though this benefit appears to diminish with increasing age 13. Similarly, in the CALGB trial, 633 patients over 65 years old were randomized to receive cyclophosphamide, methotrexate, fluorouracil (CMF), doxorubicin and cyclophosphamide (AC), or capecitabine 14. The findings showed that patients who received single-agent chemotherapy had a doubled risk of relapse or death, suggesting the advantage of the combination therapy in this age group, even though the toxicity was pronounced. In a retrospective pooled analysis of 4 randomized trials, Muss et al reported that chemotherapy reduced breast cancer mortality and recurrence rates with similar effects to younger women 6. Our findings suggest that the patients in the high-risk groups had larger tumors, more positive lymph nodes, higher grades, and HR-HER2- and inactive local therapy could obtain the improvement of BCSS after receiving chemotherapy. This conclusion is in accordance with the recommendation of the Society of Geriatric Oncology that older patients with node-positive, hormone-negative breast tumors may have the largest survival gain from chemotherapy 15.
Elderly breast cancer patients can benefit from chemotherapy but also suffer from inevitable chemotherapy toxicity. The general decline in physiological reserves and the increase in comorbidities predispose elderly women to a higher risk of toxicity. These include neuropathy, reduced left ventricular ejection fraction, congestive heart failure, myelodysplasia, acute leukemia, cardiotoxicity and secondary hematological malignancies 16–18. Another study reported that 24% of patients over 65 years old treated with docetaxel chemotherapy were hospitalized due to chemotherapy toxicity 19. In addition, anthracycline chemotherapy drugs have a higher risk of inducing myelosuppression in elderly breast cancer patients, with grade 3–4 myelosuppression occurring in 32% of patients, compared with 21% in younger patients. The difference between the two was statistically significant (P < 0.0001) 20. Therefore, the chemotherapy of elderly women should be individualized, and chemotherapy toxicity must be carefully weighed.
In the context of increased life expectancy, the treatment of elderly patients should be individualized to balance the benefits of chemotherapy and the loss of quality of life due to chemotherapy toxicity, and age should not be seen as a barrier to chemotherapy and management 21,22. In contrast, we should comprehensively evaluate the patient's general condition, cardiopulmonary and other organ functions, complications, and social support. Poor physical condition, more complications, etc., can lead to increased all-cause mortality, making adjuvant chemotherapy redundant. In conclusion, a comprehensive geriatric assessment (CGA) plays an important role in treatment decisions for elderly patients.
Although the TNM staging system is an important tool for predicting prognosis, some important prognostic factors, such as age, were not included, and the accuracy of the system's predictive results was insufficient. Our nomogram not only contains the parameters of the AJCC staging system but also includes some individual demographic and pathological characteristics and can help doctors distinguish the benefit group from chemotherapy. Therefore, it provides more comprehensiveness and convenience. However, our study also has certain limitations. First, due to limited database permissions, the information related to chemotherapy and radiation was only in the categories of "Yes" and "No/unknown". Second, the SEER database also did not provide some important prognostic-related features, including HER-2 status before 2010, Ki-67, and tumor progression, which affected the validity of our model. In addition, characteristics of geriatric assessment, such as comorbidities, physical functional status, mental health, and social support, were not included in the model. Finally, this is a retrospective study with unavoidable selection bias. To enhance the convincing power of the model, the nomogram should pass further prospective research for confirmation and supplementation.