Development and validation of a SEER-based prognostic nomogram for overall survival in elderly Asian American women with breast cancer

Introduction

Breast cancer is the most prevalent cancer and leading cause of cancer-related deaths among women, with its incidence positively correlated with age (1,2). Around half of the new breast cancer diagnoses are in women who are 60 years old or above (https://gco.iarc.fr/). However, treatment decisions for elderly breast cancer patients vary considerably (3,4). As aging is accompanied by increased fragility and comorbidities, prospective studies advocating specific treatments for the elderly remain limited due to ethical challenges, resulting in a lack of uniform treatment guidelines for this demographic (5-9).

Moreover, despite overall mortality declines, racial disparities in survival outcomes persist (10). Asian Americans are the quickest expanding racial demographic in the United States, and breast cancer is the most common malignancy among Asian American women (11-13). While population-level data consistently show that Asian American breast cancer patients as a whole have better overall and breast cancer-specific survival than non-Hispanic White patients—even after adjustment for clinicopathologic features—this aggregate advantage masks significant heterogeneity (14). Elderly Asian American women, in particular, exhibit different tumor biology, socioeconomic characteristics, and clinical outcomes compared with other populations (15,16). These differences may contribute to delayed diagnosis and poorer prognosis in certain subgroups. Given the unique challenges faced by this population, a focused examination of the survival outcomes for elderly Asian American breast cancer patients is imperative.

Numerous SEER-based prognostic nomograms for breast cancer have been developed, but most are designed for general adult populations or stratify only by age or race in isolation. No nomogram has been specifically developed and validated for elderly Asian American women—a population at the intersection of two understudied demographic dimensions. General prognostic models may have suboptimal predictive performance in this group due to differences in age-related comorbidity burden, tumor biological characteristics, treatment tolerance, and social support structures. A population-specific model can provide more precise risk stratification to support personalized clinical decision-making for this vulnerable group. Accordingly, this study aimed to develop and internally validate a prognostic nomogram for overall survival (OS) in elderly Asian American women with stage I–III breast cancer using the SEER database.

By identifying key prognostic factors, this nomogram holds potential for informing clinical decision-making and enhancing care for this population. We present this article in accordance with the TRIPOD reporting checklist (available at https://gs.amegroups.com/article/view/10.21037/gs-2026-0260/rc).

Methods

Data source and data extraction

This is a retrospective population-based prognostic model development and internal validation study using data from the SEER database [2000–2021]. Data for this study were retrieved from the Surveillance, Epidemiology, and End Results (SEER) 18 registries database, which is a publicly available, population-based cancer registry encompassing approximately 34.6% of the U.S. population. Breast cancer cases were identified using the International Classification of Diseases for Oncology, 3^rd edition (ICD-O-3) morphology and topography codes. Asian ethnicity was defined according to the SEER race/ethnicity variable, which includes self-reported Asian subgroups. The SEER database is a publicly accessible resource that provides anonymized patient data, ensuring no personally identifiable information is disclosed. Consequently, this study is exempt from requiring ethics committee approval or patient informed consent. All methodologies employed in this research strictly adhere to the guidelines established by the SEER database. The study was conducted in accordance with the Declaration of Helsinki and its subsequent amendments.

We focused on Asian female patients aged 65 years or older with a diagnosis of stage I to III breast cancer as their first and only cancer between specific years. Collected clinicopathological data included age, marital status, tumor grade, tumor-node-metastasis (TNM) stage, income, surgical history, radiotherapy, chemotherapy, molecular subtype, survival status, and survival time. Data retrieval was conducted using SEER*Stat software (version 8.4.4). The criteria for inclusion were: (I) confirmed pathological diagnosis of breast cancer; (II) primary tumor; and (III) comprehensive clinical information. The criteria for exclusion were: (I) non-Asian ethnicity; (II) incomplete tumor or treatment data; (III) a survival time shorter than 1 month; (IV) unknown age or younger than 65 years old; and (V) male breast cancer. Figure 1 illustrates the patient screening process within the SEER database.

Figure 1 Flow diagram of patient selection. BCS, breast-conserving surgery; M, metastasis; SEER, Surveillance, Epidemiology, and End Results.

Statistical analysis

Eligible patients were randomly assigned to the training cohort (70%) and validation cohort (30%) using a computer-generated random number, with stratification by TNM stage to ensure balanced distribution of disease severity. Categorical variables were compared between the two cohorts using the Pearson Chi-squared test. OS was defined as the time from diagnosis to death from any cause. Patients alive at the end of follow-up were censored at their last known follow-up date. Median follow-up was calculated using the reverse Kaplan-Meier method. Sample size adequacy was evaluated using the events-per-variable (EPV) rule for Cox proportional hazards models, with a minimum threshold of 10 events per predictor variable to minimize overfitting risk. Significant variables (P<0.05) identified through univariate analyses were subsequently included in the multivariate Cox proportional hazards regression model. Selected variables were utilized to construct the prognostic nomogram. The accuracy of the nomogram was evaluated through calibration curves, and time-dependent receiver operating characteristic (tROC) curves were employed to measure its performance. The nomogram’s discriminative ability was assessed using the concordance index (C-index), while clinical utility was evaluated via decision curve analysis (DCA). A two-tailed P<0.05 was considered statistical significance. Statistical analyses were conducted using R 4.2.1.

Results

Patient characteristics

This study included a total of 11,968 elderly Asian American breast cancer patients, split into 8,380 patients in the training cohort and 3,588 in the validation cohort. The observed number of all-cause death events in the training cohort yields an EPV ratio substantially exceeding the 10:1 recommended minimum, indicating sufficient statistical power for model development. Overfitting risk was further mitigated via 10-fold cross-validation and 1,000 bootstrap resamples. The median follow-up time was 50 months. As detailed in Table 1, baseline clinicopathological characteristics were balanced across cohorts, with no significant differences in age, grade, marital status, income, stage, T stage, N stage, subtype, surgery, radiotherapy, or chemotherapy (all P>0.05).

Identification of independent prognostic variables

Univariate Cox regression analysis indicated multiple factors significantly associated with OS (Table 2). Older age (≥75 years; hazard ratio =3.937; P<0.001), higher grade (grade III; hazard ratio =2.463; P<0.001), marital status [divorced, separated, widowed (DSW); hazard ratio =2.222, P<0.001], higher T stage (T4; hazard ratio =5.722; P<0.001), and higher N stage (N3; hazard ratio =6.178, P<0.001) were identified as risk factors for poorer OS. Additionally, hormone receptor (HR)⁻/human epidermal growth factor receptor 2 (HER2)⁻ subtype (hazard ratio =2.123; P<0.001) and mastectomy (hazard ratio =1.789; P<0.001) correlated with significantly higher mortality risk, while radiotherapy (hazard ratio =0.519; P<0.001) appeared to lower mortality risk. According to the multivariate Cox regression analysis (Table 2), these factors were confirmed as independent predictors of OS: age (age ≥75 years; hazard ratio =3.239; P<0.001), grade (grade III; hazard ratio =1.493; P<0.001), marital status (DSW; HR =1.593; P<0.001), T stage (T4; hazard ratio =2.954; P<0.001), N stage (N3; hazard ratio =4.089; P<0.001), molecular subtype (HR⁻/HER2⁻; hazard ratio =1.403; P=0.001), surgery type (mastectomy; hazard ratio =1.191; P=0.03), and radiotherapy (hazard ratio =0.567; P<0.001).

Development and verification of the prognostic nomogram

According to the independent prognostic factors identified in the multivariate analysis, a novel prognostic nomogram was developed to predict the 1-, 3-, and 5-year OS probabilities for elderly Asian American breast cancer patients (Figure 2). Each patient’s total score was calculated by summing the scores assigned to each prognostic factor, with corresponding survival probabilities subsequently determined.

Figure 2 Nomogram developed for individualized prediction of survival in elderly Asian American patients with breast cancer. BCS, breast-conserving surgery; DSW, divorced, separated, widowed; HER2, human epidermal growth factor receptor 2; HR, hormone receptor; N, node; OS, overall survival; T, tumor.

Evaluation of the nomogram’s predictive performance

The prognostic nomogram exhibited moderate-to-good discriminative ability, with C-indices of 0.779 [95% confidence interval (CI): 0.762–0.795] and 0.763 (95% CI: 0.736–0.789) in the training and validation cohorts, respectively. These results significantly surpassed those of the traditional TNM staging system, which had C-indices of 0.723 (95% CI: 0.702–0.744) and 0.719 (95% CI: 0.686–0.752) in the corresponding cohorts. Calibration plots for the 1-, 3-, and 5-year OS demonstrated excellent agreement between predicted and observed survival probabilities in both the training and validation cohorts (Figure 3A,3B). The model’s prognostic accuracy for tailored OS was further assessed with tROC curves, demonstrating superior performance compared to the conventional TNM stage in both the training and validation cohorts (Figure 3C,3D). Additionally, DCA plots illustrated that the nomogram demonstrated substantial clinical utility in predicting 1-, 3-, and 5-year OS rates (Figure 3E,3F).

Figure 3 Evaluation of the predictive performance of the prognostic model. (A) Calibration plots of the prognostic model for predicting 1-, 3-, and 5-year OS in the training cohort. (B) Calibration plots of the prognostic model for predicting 1-, 3-, and 5-year OS in the validation cohort. (C) tROC curves comparing the prognostic model with TNM stage in the training cohort. (D) tROC curves comparing the prognostic model with TNM stage in the validation cohort. (E) DCA curve analysis of the prognostic model and TNM stage in the training cohort. (F) DCA curve analysis of the prognostic model and TNM stage in the validation cohort. DCA, decision curve analysis; OS, overall survival; TNM, tumor-node-metastasis; tROC, time-dependent receiver operating characteristic.

Discussion

This study presents a prognostic nomogram for OS in elderly Asian American breast cancer patients, developed and internally validated using the SEER database. The nomogram integrates important independent prognostic factors identified through comprehensive multivariate regression analysis, including patient age, tumor grade, marital status, T stage, N stage, molecular subtypes, and treatment modalities such as radiotherapy. This intuitive predictive tool exhibits good discriminatory and calibration abilities in both training and validation cohorts, outperforming traditional TNM staging considerations. In oncologic prognostic research, a C-index of 0.70–0.80 is generally classified as moderate discriminative performance. Our nomogram achieved a validation C-index of 0.763, which falls within this range. While not outstanding, this represents a meaningful improvement over the conventional TNM staging system (C-index =0.719), as it integrates additional prognostic factors including age, tumor grade, marital status, molecular subtype, and treatment modalities to provide more granular risk stratification than anatomic staging alone.

We have reinforced existing literature linking higher age at diagnosis to poorer prognostic outcomes, attributable to treatment-related complications, the presence of comorbidities, and variances in tumor biology in elderly patients (17-20). Additionally, higher tumor grades, significant lymph node invasion, and unfavorable molecular subtypes (such as triple-negative breast cancer) are recognized as adverse prognostic indicators across various racial and ethnic groups (21-23). This nomogram may serve as an adjunctive tool to support shared decision-making and risk stratification in elderly Asian American breast cancer patients, but prospective and external validation are required before it can be recommended for widespread clinical implementation.

Notably, marital status validates its role as an independent prognostic predictor, as unmarried patients experience significantly worse survival outcomes compared to their married counterparts (24-26). This aligns with prior research emphasizing the significance of social support and family dynamics in tumor prognosis; married individuals may fare better due to enhanced treatment adherence, reduced psychological stress, and improved access to medical resources (25,27-30). Incorporating this sociodemographic factor into predictive models indicates that cancer prognosis transcends mere clinical and tumor characteristics, encompassing multiple contributing factors. Multimodal cancer treatments, including surgery and radiotherapy, likewise influence survival outcomes among our elderly Asian American breast cancer sample (31-37). Although further investigation into the biological interactions between treatments and specific tumor characteristics is warranted, our results underscore the essentiality of integrating treatment details into prognostic model development. Formulating personalized, risk-adapted treatment strategies for high-risk elderly Asian American breast cancer patients identified by our nomogram is crucial.

The robust calibration, tROC curves, and DCAs collectively confirm the nomogram’s good predictive accuracy. Clinicians may leverage this prognostic tool to support risk prediction, set appropriate patient expectations, guide individualized treatment planning, and facilitate shared decision-making with elderly Asian American breast cancer patients. The nomogram’s incremental performance over the conventional TNM staging system underscores the value of developing population-specific risk assessment models that account for the nuanced, intersectional factors shaping health outcomes. Neglecting to disaggregate data and research findings by race, age, and other key social determinants can obscure critical disparities and hinder the delivery of equitable, high-quality care (18,22,38-40). The present study serves as a compelling example of how targeted prognostic models can advance precision oncology and promote the realization of health equity.

Ultimately, the clinical significance of this prognostic nomogram lies in its potential to address the historical underrepresentation and healthcare disparities faced by elderly Asian American breast cancer patients. This nomogram may be used as an adjunct to facilitate shared decision-making and risk stratification in elderly Asian American patients with breast cancer. However, prospective studies and external validation are necessary before its widespread clinical application can be recommended.

This study acknowledges inherent limitations, including the retrospective nature, which carries selection bias risk. First, this study performed only split-sample internal validation within the SEER database. Internal validation tends to yield overoptimistic performance estimates, and the generalizability of the nomogram to non-SEER populations, other healthcare systems, or different geographic regions remains unconfirmed. External validation in independent, geographically distinct cohorts is a critical next step to verify model performance before clinical application. Second, this study combines all Asian American subgroups into a single analytical group, which introduces potential aggregation bias. Asian American is not a biologically or socially homogeneous category: Chinese, Japanese, Filipina, Korean, Vietnamese, and South Asian women differ in tumor molecular profiles, treatment preferences, socioeconomic status, immigration history, and survival outcomes. Our model may not perform equally well across all Asian subgroups. Future studies with sufficient sample sizes for disaggregated analyses are needed to develop subgroup-specific prognostic tools and understand within-Asian disparities. Additionally, the SEER database lacks specific clinical variables, such as performance status, comorbidities, and detailed systemic treatment information, as well as individual-level social determinants (language, health literacy, acculturation, immigration history, and detailed insurance information). Future prospective studies incorporating a broader set of predictive factors can enhance the nomogram’s accuracy. Furthermore, continued refinement and external validation of this tool in diverse healthcare settings will be crucial to realizing its transformative impact on breast cancer care and advancing the goal of health equity.

Conclusions

The proposed nomogram has the potential to support shared decision-making and improve risk stratification for elderly Asian American breast cancer patients. Nevertheless, further prospective and external validation is needed prior to broad clinical adoption.

Acknowledgments

None.

Footnote

Reporting Checklist: The authors have completed the TRIPOD reporting checklist. Available at https://gs.amegroups.com/article/view/10.21037/gs-2026-0260/rc

Peer Review File: Available at https://gs.amegroups.com/article/view/10.21037/gs-2026-0260/prf

Funding: This work was supported by grants from the National Natural Science Foundation of China (No. 82503927), the Guangdong Basic and Applied Basic Research Foundation (No. 2026A1515012552), the Fundamental Research Funds for the Central Universities (No. YG2024QNA06), and the Supporting Research Fund for High-Level Full-Time Talents Introduction of Guangdong Provincial People’s Hospital (No. KY012025021).

Conflicts of Interest: All authors have completed the ICMJE uniform disclosure form (available at https://gs.amegroups.com/article/view/10.21037/gs-2026-0260/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 and its subsequent amendments.

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. Bray F, Laversanne M, Sung H, et al. Global cancer statistics 2022: GLOBOCAN estimates of incidence and mortality worldwide for 36 cancers in 185 countries. CA Cancer J Clin 2024;74:229-63. [Crossref] [PubMed]
2. Nasrazadani A, Marti JLG, Kip KE, et al. Breast cancer mortality as a function of age. Aging (Albany NY) 2022;14:1186-99. [Crossref] [PubMed]
3. Jenkins EO, Deal AM, Anders CK, et al. Age-specific changes in intrinsic breast cancer subtypes: a focus on older women. Oncologist 2014;19:1076-83. [Crossref] [PubMed]
4. Petkov VI, Miller DP, Howlader N, et al. Breast-cancer-specific mortality in patients treated based on the 21-gene assay: a SEER population-based study. NPJ Breast Cancer 2016;2:16017. [Crossref] [PubMed]
5. Patnaik JL, Byers T, Diguiseppi C, et al. The influence of comorbidities on overall survival among older women diagnosed with breast cancer. J Natl Cancer Inst 2011;103:1101-11. [Crossref] [PubMed]
6. Parks RM, Holmes HM, Cheung KL. Current Challenges Faced by Cancer Clinical Trials in Addressing the Problem of Under-Representation of Older Adults: A Narrative Review. Oncol Ther 2021;9:55-67. [Crossref] [PubMed]
7. Varghese F, Wong J. Breast Cancer in the Elderly. Surg Clin North Am 2018;98:819-33. [Crossref] [PubMed]
8. Baban CK, Devane L, Geraghty J. Change of paradigm in treating elderly with breast cancer: are we undertreating elderly patients? Ir J Med Sci 2019;188:379-88. [Crossref] [PubMed]
9. Freedman RA, Keating NL, Lin NU, et al. Breast cancer-specific survival by age: Worse outcomes for the oldest patients. Cancer 2018;124:2184-91. [Crossref] [PubMed]
10. Yedjou CG, Sims JN, Miele L, et al. Health and Racial Disparity in Breast Cancer. Adv Exp Med Biol 2019;1152:31-49. [Crossref] [PubMed]
11. Truong A, McKinley M, Gomez SL, et al. The role of ethnic enclaves and neighborhood socioeconomic status in invasive breast cancer incidence rates among Asian American, Native Hawaiian, and Pacific Islander females in California. Cancer Causes Control 2025;36:183-9. [Crossref] [PubMed]
12. Medina HN, Callahan KE, Morris CR, et al. Cancer Mortality Disparities among Asian American and Native Hawaiian/Pacific Islander Populations in California. Cancer Epidemiol Biomarkers Prev 2021;30:1387-96. [Crossref] [PubMed]
13. Eden CM, Johnson J, Syrnioti G, et al. The Landmark Series: The Breast Cancer Burden of the Asian American Population and the Need for Disaggregated Data. Ann Surg Oncol 2023;30:2121-7. [Crossref] [PubMed]
14. Li Y, Li D, Tao X, et al. Differences in Survival and Associated Characteristics Between Asian American and White Patients With Breast Cancer. Clin Breast Cancer 2026;26:1-8. [Crossref] [PubMed]
15. Eden CM, Jao L, Syrnioti G, et al. Breast cancer in women of Asian heritage: disparity trends in the Asian American breast cancer population literature. Current Breast Cancer Reports 2024;16:351-8.
16. Thaploo A, Kohli K, Wang S, et al. Disparities in Stage at Presentation for Disaggregated Asian American, Native Hawaiian, and Pacific Islander Patients with Breast Cancer. Ann Surg Oncol 2025;32:3317-30. [Crossref] [PubMed]
17. Sung WWY, Sharma K, Chan AW, et al. A narrative review of the challenges and impact of breast cancer treatment in older adults beyond cancer diagnosis. Ann Palliat Med 2024;13:1521-9. [Crossref] [PubMed]
18. Trapani D. Adjuvant Chemotherapy in Older Women With Early Breast Cancer. J Clin Oncol 2023;41:1652-8. [Crossref] [PubMed]
19. Schonberg MA, Marcantonio ER, Li D, et al. Breast cancer among the oldest old: tumor characteristics, treatment choices, and survival. J Clin Oncol 2010;28:2038-45. [Crossref] [PubMed]
20. Di Lascio S, Tognazzo E, Bigiotti S, et al. Breast cancer in the oldest old (≥ 89 years): Tumor characteristics, treatment choices, clinical outcomes and literature review. Eur J Surg Oncol 2021;47:796-803. [Crossref] [PubMed]
21. Loibl S, Poortmans P, Morrow M, et al. Breast cancer. Lancet 2021;397:1750-69. [Crossref] [PubMed]
22. Nolan E, Lindeman GJ, Visvader JE. Deciphering breast cancer: from biology to the clinic. Cell 2023;186:1708-28. [Crossref] [PubMed]
23. Harbeck N, Penault-Llorca F, Cortes J, et al. Breast cancer. Nat Rev Dis Primers 2019;5:66. [Crossref] [PubMed]
24. Jiao D, Ma Y, Zhu J, et al. Impact of Marital Status on Prognosis of Patients With Invasive Breast Cancer: A Population-Based Study Using SEER Database. Front Oncol 2022;12:913929. [Crossref] [PubMed]
25. Ding W, Ruan G, Lin Y, et al. Dynamic changes in marital status and survival in women with breast cancer: a population-based study. Sci Rep 2021;11:5421. [Crossref] [PubMed]
26. Zhai Z, Zhang F, Zheng Y, et al. Effects of marital status on breast cancer survival by age, race, and hormone receptor status: A population-based Study. Cancer Med 2019;8:4906-17. [Crossref] [PubMed]
27. Osborne C, Ostir GV, Du X, et al. The influence of marital status on the stage at diagnosis, treatment, and survival of older women with breast cancer. Breast Cancer Res Treat 2005;93:41-7. [Crossref] [PubMed]
28. Paranjpe A, Zheng C, Chagpar AB. Disparities in Breast Cancer Screening Between Caucasian and Asian American Women. J Surg Res 2022;277:110-5. [Crossref] [PubMed]
29. Taparra K, Qu V, Pollom E. Disparities in Survival and Comorbidity Burden Between Asian and Native Hawaiian and Other Pacific Islander Patients With Cancer. JAMA Netw Open 2022;5:e2226327. [Crossref] [PubMed]
30. Freeman JQ, Li JL, Fisher SG, et al. Declination of Treatment, Racial and Ethnic Disparity, and Overall Survival in US Patients With Breast Cancer. JAMA Netw Open 2024;7:e249449. [Crossref] [PubMed]
31. Swaminathan V, Spiliopoulos MK, Audisio RA. Choices in surgery for older women with breast cancer. Breast Care (Basel) 2012;7:445-51. [Crossref] [PubMed]
32. 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;21:786-94. [Crossref] [PubMed]
33. Frebault J, Bergom C, Kong AL. Surgery in the Older Patient with Breast Cancer. Curr Oncol Rep 2019;21:69. [Crossref] [PubMed]
34. Han T, Shi M, Chen Q, et al. Effect of adjuvant radiotherapy after breast-conserving surgery in elder women with early-stage breast cancer: a propensity-score matching analysis. Front Oncol 2023;13:1012139. [Crossref] [PubMed]
35. Walker GA, Kaidar-Person O, Kuten A, et al. Radiotherapy as sole adjuvant treatment for older patients with low-risk breast cancer. Breast 2012;21:629-34. [Crossref] [PubMed]
36. Xie W, Cao M, Zhong Z, et al. Survival outcomes for breast conserving surgery versus mastectomy among elderly women with breast cancer. Breast Cancer Res Treat 2022;196:67-74. [Crossref] [PubMed]
37. Morgan JL, George J, Holmes G, et al. Breast cancer surgery in older women: outcomes of the Bridging Age Gap in Breast Cancer study. Br J Surg 2020;107:1468-79. [Crossref] [PubMed]
38. Ko NY, Hong S, Winn RA, et al. Association of Insurance Status and Racial Disparities With the Detection of Early-Stage Breast Cancer. JAMA Oncol 2020;6:385-92. [Crossref] [PubMed]
39. Chen HL, Zhou MQ, Tian W, et al. Effect of Age on Breast Cancer Patient Prognoses: A Population-Based Study Using the SEER 18 Database. PLoS One 2016;11:e0165409. [Crossref] [PubMed]
40. Moran MS, Ho AY. Radiation Therapy for Low-Risk Breast Cancer: Whole, Partial, or None? J Clin Oncol 2022;40:4166-72. [Crossref] [PubMed]